Download E1.1 Circuit Analysis Problem Sheet 1

Ver 3537
E1.1 Analysis of Circuits (2014)
E1.1 Circuit Analysis
Problem Sheet 1 (Lectures 1 & 2)
Key: [A]= easy ... [E]=hard
1. [A] One of the following circuits is a series circuit and the other is a parallel circuit. Explain which
is which.
(a)
(b)
2. [B] Find the power absorbed by by each of the subcircuits A and B given that the voltage and
current are 10 V and 2 A as shown.
3. [B] For each of the four circuits below, find the power absorbed by the voltage source (PV ), the
power absorbed by the current source (PI ) and the total power absorbed (PV + PI ).
(a)
(b)
(c)
(d)
4. [B] Determine the voltage VX in the following circuit.
5. [B] Determine the current IX in the following circuit.
6. [B] What single resistor is equivalent to the three resistor sub-circuit shown below?
Problem Sheet 1
Page 1 of 2
Ver 3537
E1.1 Analysis of Circuits (2014)
7. [B] What single resistor is equivalent to the three resistor sub-circuit shown below?
8. [C] What single resistor is equivalent to the five resistor sub-circuit shown below?
9. [A] If a resistor has a conductance of 8 µS, what is its resistance?
10. [B] Determine the voltage across each of the resistors in the following circuit and the power dissipated
in each of them. Calculate the power supplied by the voltage source.
11. [B] Determine the current through each of the resistors in the following circuit and the power dissipated in each of them. Calculate the power supplied by the current source.
12. [B] Determine R1 so that Y = 14 X.
13. [B] Choose R1 and R2 so that Y = 0.1X and R1 + R2 = 10 MΩ.
14. [D] You have a supply of resistors that have the values {10, 12, 15, 18, 22, 27, 33, 39, 47, 56, 68, 82} ×
10n Ω for all integer values of n. Thus, for example, a resistor of 390 Ω is available and the next higher
value is 470 Ω. Show how, by combining two resistors in each case, it is possible to make networks
whose equivalent resistance is (a) 3 kΩ, (b) 4 kΩ and (c) as close as possible to 3.5 kΩ. Determine is
the worst case percentage error that might arise if, instead of combining resistors, you just pick the
closest one available.
Problem Sheet 1
Page 2 of 2
Ver 2427
E1.1 Analysis of Circuits (2014)
E1.1 Circuit Analysis
Problem Sheet 1 - Solutions
1. Circuit (a) is a parallel circuit: there are only two nodes and all four components are connected
between them.
Circuit (b) is a series circuit: each node is connected to exactly two components and the same current
must flow through each.
2. For subcircuit B the voltage and current correspond to the passive sign convention (i.e. the current
arrow in the opposite direction to the voltage arrow) and so the power absorbed by B is given by
V × I = 20 W.
For device A we need to reverse the direction of the current to conform to the passive sign convention.
Therefore the power absorbed by A is V × −I = −20 W.
As must always be true, the total power absorbed by all components is zero.
3. The power absorbed is positive if the voltage and current arrows go in opposite directions and negative
if they go in the same direction. So we get: (a) PV = +4, PI = −4, (b) PV = +4, PI = −4, (c)
PV = −4, PI = +4, (a) PV = −4, PI = +4. In all cases, the total power absorbed is PV + PI = 0.
4. We can find a path (shown highlighted below) from the bottom to the top of the VX arrow that
passes only through voltage sources and so we just add these up to get the total potential difference:
VX = (−3) + (+2) + (+9) = +8 V.
5. If we add up the currents flowing out of the region shown highlighted below, we obtain IX −5−1+2 =
0. Hence IX = 4 A.
6. The three series resistors are equivalent to a single resistor with a value of 1 + 5 + 2 = 8 kΩ.
7. The three series resistors are equivalent to a single resistor with a value of
0.588 kΩ.
1
1/1+1/5+1/2
=
1
1.7
=
8. We can first combine the parallel 2 k and 3 k resistors to give 2×3
2+3 = 1.2 k. This is then in series with
the 4 k resistor which makes 5.2 k in all. Now we just have three resistors in parallel to give a total
1
= 0.718kΩ.
of 1/1+1/15+1/5.2 = 1.39
9. The resistance is
Solution Sheet 1
1
8×10−6
= 125 kΩ
Page 1 of 2
Ver 2427
E1.1 Analysis of Circuits (2014)
10. [Method 1]: The resistors are in series and so form a potential divider. The total series resistance is
7 k, so the voltages across the three resistors are 14 × 17 = 2 V, 14 × 72 = 4 V and 14 × 47 = 8 V. The
2
2
2
power dissipated in a resistor is VR , so for the three resistors, this gives 21 = 4 mW, 42 = 8 mW and
82
4 = 16 mW.
[Method 2]: The total resistance is is 7 k so the current flowing in the circuit is 14
7 = 2 mA. The voltage
across a resistor is IR which, in for these resistors, gives 2 × 1 = 2 V, 2 × 2 = 4 V and 2 × 4 = 8 V.
The power dissipated is V I which gives 2 × 2 = 4 mW, 4 × 2 = 8 mW and 8 × 2 = 16 mW. The
current through the voltage source is 2 mA, so the power it is supplying is V I = 14 × 2 = 28 mW.
This is, inevitably, equal to the sum of the power disspipated by the three resistors: 4 + 8 + 16 = 28.
11. [Method 1]: The resistors are in parallel and so form a current divider: the 21 mA will divide in
proportion to the conductances: 1 mS, 0.5 mS and 0.25 mS. The total conductance is 1.75 mS, so the
0.5
1
= 12 mA, 21 × 1.75
= 6 mA and 21 × 0.25
three resistor currents are 21 × 1.75
1.75 = 3 mA. The power
2
2
dissipated in a resistor is I R which gives 12 × 1 = 144 mW, 62 × 2 = 72 mW and 32 × 4 = 36 mW.
[Method 2]: The equivalent resistance of the three resistors is 1/1+1/12+1/4 = 47 kΩ. Therefore the
voltage across all components in the parallel circuit is 21 × 74 = 12 V. The current through a resistor
12
12
is VR which gives 12
1 = 12 mA, 2 = 6 mA and 4 = 3 mA. The power dissipated in a resistor is V I
which gives 12 × 12 = 144 mW, 12 × 6 = 72 mW and 12 × 3 = 36 mW. The power supplied by the
current source is 12 × 21 = 252 mW which as expected equals 144 + 72 + 36.
12. The resistors form a potential divider, so
R1 = 12 k.
Y
X
=
4
R1 +4 .
So we want
4
R1 +4
=
1
4
⇒ R1 + 4 = 16 ⇒
Y
1
2
2
13. The resistors form a potential divider, so X
. So we want R1R+R
= 10
= R1R+R
and R1 + R2 =
2
2
10 MΩ. Substituting one into the other and cross-multiplying gives 10R2 = 10 MΩ ⇒ R2 = 1 MΩ.
Substituting this into the simpler of the two initial equations gives R1 = 10 − 1 = 9 MΩ.
14. (a) 3 k = 1.5 k + 1.5 k = 3.3 k||33 k, (b) 4 k = 3.9 k + 100, (c) 3.488 k = 3.9 k||33 k.
To make an exhaustive search for creating a resistance of R, you need to consider two possibilities:
(i) for two resistors in series, the largest of the two resistors must be in the range [ 12 R, R] or (ii) for
two resistors in parallel, the smallest resistor must be in the range [R, 2R]. In both cases there are
at most four possibilities, so you need to consider up to eight possibilities in all. So, for example, for
R = 3.5 k, we would consider the following possibilities: (i) 1.5 k + 1.8 k = 3.3 k, 1.8 k + 1.8 k = 3.6 k,
2.2 k + 1.2 k = 3.4 k, 2.7 k + 0.82 k = 3.52 k and (ii) 3.9 k||33 k = 3.488 k, 4.7 k||15 k = 3.579 k,
5.6 k||10 k = 3.59 k, 6.8 k||6.8 k = 3.4 k. The choice with least error is the one given above.
Since we are interested in % errors, we need to consider the ratio between resistor values. The largest
15
ratio between successive resistors is the series is 12
= 1.25 (this includes the wraparound ratio of
100
=
1.22).
The
worst-case
percentage
error
will
arise
if our target resistance is the mean of these
82
1.5
two values, 13.5. The percentage error in choosing either one is then 13.5
= 11.1%.
Solution Sheet 1
Page 2 of 2
Ver 2427
E1.1 Analysis of Circuits (2015)
E1.1 Circuit Analysis
Problem Sheet 2 (Lectures 3 & 4)
Key: [A]= easy ... [E]=hard
1. [B] Calculate VX and IX in the following circuit using (a) nodal analysis and (b) simplifying the
circuit by combining parallel resistors.
2. [B] Calculate VX and IX in the following circuit using (a) nodal analysis and (b) simplifying the
circuit by combining parallel resistors.
3. [C] Calculate VX in the following circuit using (a) nodal analysis and (b) superposition.
4. [C] Calculate VX in the following circuit.
5. [C] Calculate VX in the following circuit.
6. [C] Calculate VX in the following circuit.
Problem Sheet 2
Page 1 of 3
Ver 2427
E1.1 Analysis of Circuits (2015)
7. [C] Calculate VX in the following circuit. The value of the dependent current source is 99 time the
current flowing through the 1 V voltage source.
8. [C] In the following circuit calculate VX in terms of V and I using (a) nodal analysis and (b)
superposition.
9. [C] Calculate VX and IX in the following circuit using (a) nodal analysis and (b) superposition.
10. [C] Determine an expression for IX in terms of V in the following circuit. Determine the value of V
that will make IX = 0.
11. [C] Calculate VX in the following circuit using (a) nodal analysis and (b) superposition.
12. [C] Calculate VX in the following circuit which includes a dependent voltage source.
Problem Sheet 2
Page 2 of 3
Ver 2427
E1.1 Analysis of Circuits (2015)
13. [C] Find the equivalent resistance of the network shown below.
14. [D] Prove that if VAB = 0 , then R = 4 kΩ in the following circuit. The circuit is used to detect
small changes in R from its nominal value of 4 kΩ. Find an expression for VAB as a function of R.
If changes in VAB of 10 mV can be detected, what is the smallest detectable change in R .
15. [D] Calculate VX in the following circuit. You can either use nodal analysis directly or else simplify
the circuit a little to reduce the number of nodes.
16. [D] Calculate VX in the following circuit which includes a floating dependent voltage source.
Problem Sheet 2
Page 3 of 3
Ver 6186
E1.1 Analysis of Circuits (2015)
E1.1 Circuit Analysis
Problem Sheet 2 - Solutions
Note: In many of the solutions below I have written the voltage at node X as the variable X instead of
VX in order to save writing so many subscripts.
1. [Nodal analysis] KCL at node VX gives
which VX = 8.
VX −14
1
+ V2X + V4X = 0 which simplifies to 7VX − 56 = 0 from
[Parallel resistors] We can merge the 2 Ω and 4 Ω resistors to make one of
1.33
Now we have a potential divider, so VX = 14 × 2.33
= 8 V.
2×4
2+4
=
4
3
Ω as shown below.
In both cases, we can now calculate IX = V4X = 2 A. Note that when we merge the two resistors, IX
is no longer a distinct current on the diagram.
Original
Simplified
2. [Nodal Analysis] KCL at node Vx gives 5 +
which VX = −4.
VX
1
+
VX
4
= 0 which simplifies to 20 + 5VX = 0 from
[Parallel Resistors] We can combine the 1 Ω and 4 Ω resistors to make one of
below. Now we have VX = −5 × 0.8 = −4 V.
Original
In both cases, we can now calculate IX =
VX
4
1×4
1+4
=
4
5
Ω as shown
Simplified
= −1 A.
In this question, you have to be a bit careful about the sign used to represent currents. Whenever you
use Ohm’s law, you must be sure that you use the passive sign convention (with the current arrow
in the opposite direction to the voltage arrow); this is why the current through the 0.8 Ω resistor is
−5 A rather than +5 A.
Solution Sheet 2
Page 1 of 5
Ver 6186
E1.1 Analysis of Circuits (2015)
3. [Nodal Analysis] KCL at node Vx gives
which VX = −2.
VX −6
4
+
VX
1
+ 4 = 0 which simplifies to 5VX + 10 = 0 from
[Superposition] (i) If we set the current source to zero, then the 4 Ω resistor connected to it plays
no part in the circuit and we have a potential divider giving VX = 6 × 15 = 1.2. (ii) We now set the
voltage source to zero and then simplify the resultant circuit as shown below. Being careful with
signs, we now get VX = −4 × 0.8 = −3.2. Adding these two values together gives a final answer of
VX = −2.
Original
I=0
V =0
V = 0 simplified
4. We first label the nodes; we only need two variables because of the floating voltage source. KCL
−13)
+ X−(Y
+ X−Y
= 0 which gives 11X − 9Y = 1. KCL at the supernode
at X gives X−20
30
20
10
−13)
Y −20
Y −X
{Y, Y − 13}gives 15 + 10 + (Y −13)−X
+ (Y 10
= 0 which gives −9X + 19Y = 197. Solving
20
these two simultaneous equations gives X = 14 and Y = 17.
5. We first label the nodes; there are only two whose voltage is unknown. Working in mA and kΩ, KCL
Y −X
Y
Y −60
at X gives X−240
+ X−Y
3
6 + 10 = 0 which gives 3X − Y = 420. KCL at Y gives
6 + 24 + 12 = 0
which gives −4X + 7Y = 120. Solving these simultaneous equations gives X = 180 and Y = 120.
6. We first label the nodes; since the two nodes having unknown voltages are joined by a fixed voltage
source, we only need one variable. We write down KCL for the supernode {X, X − 50} (shaded in
X
the diagram) which gives (X−50)−300
+ (X−50)
+ X−300
+ 90
= 0 which simplifies to 20X = 3500 or
90
10
10
X = 175 V.
Solution Sheet 2
Page 2 of 5
Ver 6186
E1.1 Analysis of Circuits (2015)
7. There is only one node with an unknown voltage, namely X. However, there is a dependent current
source, so we need to express its value in terms of the node voltages: 99I = 99 × X−1
125 where we are
X−1
X−1
expressing currents in mA. So now we can apply KCL to node X to obtain X−10
+
1
125 +99× 125 = 0
which simplifies to 225X = 1350 from which X = 6.
8. [Nodal Analysis] Using KCL at node X gives
1
V + 20
X = 21
21 I.
X−V
20
+
X
1
− I = 0 which we can rearrange to give
1
[Superposition] If we set I = 0 then we have a voltage divider in which X = 21
V (see middle
diagram). If we set V = 0 then (see right diagram) we can combine the two parallel resistors as
20×1
20
20
20+1 = 21 Ω and it follows that X = 21 I. By superposition, we can add these two expression
20
1
V + 21
I.
together to give X = 21
Original
I=0
V =0
9. [Nodal Analysis] We can easily see that the 4 A current flowing through the leftmost resistor means
the top left node has a voltage of −8 (although actually we do not need to calculate this because of the
X
isolating effect of the current source). Using KCL at the supernode {X−4, X} gives −4+ X−4
2 + 2 =0
6
which we can rearrange to give 2X = 12 or X = 6. It follows that IX = 2 = 3 A.
[Superposition] If we set I = 0 then we have a voltage divider (since the resistors are in series) in
2
which X = 4 × 2×2
= 2 and IX = 1 A (see middle diagram). If we set V = 0 then (see right diagram)
we can combine the two parallel resistors as 2×2
2+2 = 1 Ω and it follows that X = 4 and, by current
division, that IX = 2 A. By superposition, we can add these two expression together to give X = 6 V
and IX = 3 A.
Original
Solution Sheet 2
I=0
V =0
Page 3 of 5
Ver 6186
E1.1 Analysis of Circuits (2015)
X−V
10. [Nodal Analysis] We first label the unknown node as X. Now, KCL at this node gives −6+ X
=
3 + 6
1
X
1
0 which rearranges to give 3X = V + 36 from which X = 3 V + 12. Now IX = 3 so IX = 9 V + 4.
[Superposition] If we set I = 0 (see middle diagram) then IX = V9 . If we set V = 0 then (see
right diagram) we have a current divider in which the 6 A current divides in proportion to the
1/3
2
1
conductances. So IX = 6 × 1/3+
1/6 = 6 × 3 = 4. Adding these results together gives IX = 9 V + 4.
If V = −36 then IX = 0.
Original
11. [Nodal Analysis] KCL at node X gives
from which X = 3.
I=0
X−10
2
+
X
2
+
V =0
X−(−3)
3
= 0 which rearranges to give 8X = 24
[Superposition] If we set the left source to zero (see middle diagram) then, the two parallel 2 Ω
resistors are equivalent to 1 Ω and so we have a potential divider and X = −3 × 14 = −0.75. If we
set the other source to zero (see right diagram) we can combine the 2 Ω and 3 Ω parallel resistors to
1.2
obtain 2×3
2+3 = 1.2 Ω. We again have a potential divider giving X = 10 × 2+1.2 = 3.75. Adding these
together gives X = −0.75 + 3.75 = 3 V.
Original
U1 = 0
U2 = 0
12. KCL at node Y gives Y 1−4 + Y −X
= 0 which rearranges to −X +6Y = 20. We also have the equation
5
of the dependent voltage source: X = −6Y . We can conveninetly eliminate 6Y between these two
to give −2X = 20 and so X = −10.
13. We first pick a ground reference at one end of the network and label all the other nodes. The
equivalent resistance is now VI . We assume that we know V and then calculate I. KCL at node
A gives A−V
+ A−B
+ A5 = 0 from which 3A − B = V or B = 3A − V . KCL at node B gives
5
5
B−V
B−A
B
+ 5 = 0 from which 11B − 5A = V . Substituting B = 3A − V into this equation gives
25 +
5
8
33A − 5A = 12V or A = 12
28 V which in turn gives B = 3A − V = 28 V . The current I is the sum of
the currents throught the rightmost two 5 Ωresistors: I = A5 + B5 = 71 V . So the equivalent resistance
is VI = 7 Ω.
Solution Sheet 2
Page 4 of 5
Ver 6186
E1.1 Analysis of Circuits (2015)
14. If VAB = 0 then no current flows through the 2 k resistor, so the two vertical resistor chains form
4
= 50 V. Since VAB = 0, VB = VA = 50 =
potential dividers. In the leftmost chain, VA = 100 × 4+4
R
100 × 4+R which implies that R = 4 k.
KCL at nodes A gives A−100
+ A4 + A−B
= 0 which gives 4A − 2B = 100 or B = 2A − 50. KCL at
4
2
B−100
B
B−A
node B now gives
+
+
=
0
into which we can substitute the expression for B to get
4
R
2
2A−150
2A−50
A−50
+ R + 2 = 0 from which A = 100+125R
4
4+2R . Substituting this into B = 2A − 50 gives
150R
B = 4+2R and hence A − B = 100−25R
.
If
we
can
detect a value of A − B = 10 mV = 0.01 then the
4+2R
corresponding value of R is the solution to 100−25R
4+2R = 0.01 which gives 25.02R = 99.96 from which
R = 3995.2 Ω which is a change of 4.8 Ω or 0.12%.
15. We label the nodes as shown below (using X instead of VX for ease of writing). Note that when
we have labelled the upper node of the floating voltage source as Y we can label the lower node as
X−Y
+ X−Z
= 0 which gives
Y + 13 and do not need another variable. KCL at node X gives X−19
2
3 + 4
Z
Z−X
Z−Y −15
13X −3Y −4Z = 114. KCL at node Z gives 2 + 3 +
= 0 which gives −2X −3Y +8Z = 45.
2
Y +15−Z
Finally, KCL at the supernode {Y, Y + 15} gives Y −X
+
= 0 from which X − 3Y + 2Z = 30.
4
2
Solving these three simultaneous equations gives X = 11, Y = −1 and Z = 8.
Original
Simplified
An alternative approach is to notice that the three rightmost components are in series and so you
can reorder them without affecting the rest of the circuit to give the simplified circuit shown above.
Now, we only have two unknowns and hence only two simultaneous equations to solve. KCL at node
X gives X−19
+ X−W3 −15 + X−W
= 0 which gives 6X −3W = 87. KCL at the supernode {W, W +15}
2
6
W −X
W +15
gives 6 + W +15−X
+
=
0 from which 3X − 6W = 75. These equations are easily solved to
3
2
give X = 11 and W = −7.
16. Since the floating voltage source is a dependent voltage source, we need to label its two ends with
separate variables (see below). We now write down a KCL equation for the supernode shown shaded:
Y −48
+ Y4 + X−48
+X
4
9
6 = 0 which simplifies to 10X + 18Y = 624.
4
We also need to express the voltage source value in terms of nodal voltages: Y −X = 8I = 8× X
6 = 3X
7
7
which rearranges to give Y = 3 X. Substituting this in the previous equation gives 10X + 18 × 3 X =
624 which simplifies to 52X = 624 from which X = 12.
Solution Sheet 2
Page 5 of 5
Ver 2379
E1.1 Analysis of Circuits (2014)
E1.1 Circuit Analysis
Problem Sheet 3 (Lectures 5, 6, 7 & 8)
Key: [A]= easy ... [E]=hard
1. [B] Calculate the Thévenin and Norton equivalent networks at the terminals A and B for each of
the following.
(a)
(b)
2. [B] Use nodal analysis to calculate an expression for A in Fig. 2 in terms of I and then rearrange
this to give I in terms of A. Show how these expressions are related to the Thévenin and Norton
equivalent networks at the terminals A and B.
Fig. 2
Fig. 3
Fig. 4
3. [B] Determine X in Fig. 3 when (a) U = +5 V and (b) U = −5 V. Assume that the diode has a
forward voltage drop of 0.7 V.
4. [B] In Fig. 4, calculate I and the power dissipation in the resistor and in the diode. Assume that
the diode has a forward voltage drop of 0.7 V.
5. [B] Find the gains
X
U
and
Y
U
in the following circuits:
(a)
(b)
6. [C] Calculate the Thévenin and Norton equivalent networks at the terminals A and B in Fig. 6 in
two ways (a) by combining resistors to simplify the circuit and (b) by using nodal analysis to express
A in terms of I.
Fig. 6
Problem Sheet 3
Page 1 of 5
Ver 2379
E1.1 Analysis of Circuits (2014)
7. [C] Find the current I in Fig. 7 in two ways: (a) by nodal analysis and (b) by combining the leftmost
three components into their Thévenin equivalent.
Fig. 7
Fig. 8
8. [C] For what value of R in Fig. 8 will the power dissipation in R be maximized. Find the power
dissipation in R in this case.
9. [C] Find the Thévenin equivalent of the circuit between nodes A and B in two ways: (a) by performing
a sequence of Norton↔Thévenin transformations and (b) using superposition to find the open-circuit
voltage and combining resistors to find the Thévenin resistance.
10. [C] State whether the feedback in the following circuits is positive or negative:
(a)
11. [C] Find the gain
Problem Sheet 3
(b)
X
U
(c)
(d)
in the following circuit:
Page 2 of 5
Ver 2379
E1.1 Analysis of Circuits (2014)
12. [C] Find expressions for X, Y and Z in terms of U1 , U2 and U3 .
(a)
(b)
(c)
13. [C] Choose values of R1 and R2 in Fig. 13 so that X = 2U2 − 3U1 .
Fig. 13
Fig. 14
14. [C] Find an expression for Y in Fig. 14 in terms of U1 and U2 .
15. [C] In the circuit diagram, the potentiometer resistance between the slider and ground is a × 40 kΩ
where 0 ≤ a ≤ 1. Find the gain of the circuit, X
U as a function of a. What is the range of gains that
the circuit can generate as a is varied.
16. [C] By replacing the rightmost three components in Fig. 16 by their Thévenin equivalent, find X
when (a) U = 0 V and (b) U = 5 V. Assume that the diode has a forward voltage drop of 0.7 V.
Determine the value of U at which the diode switches between operating regions.
Fig. 16
Fig. 18
17. [C] In the block diagram of Fig. 17, F, G, H represent the gains of the blocks. Find the overall gain
Y
X.
Fig. 17
Problem Sheet 3
Page 3 of 5
Ver 2379
E1.1 Analysis of Circuits (2014)
18. [D] The circuit of Fig. 18 includes a dependent current source whose value is proportional to the
current J. Find the Thévenin equivalent of the circuit by two methods: (a) use nodal analysis
including the current I and express the voltage VAB as a function of I and (b) assume I = 0 and
find (i) the open circuit voltage, VAB as a function of U and (ii) the short-circuit current (with A
joined to B) as a function of U . Hence find the Thévenin equivalent of the circuit.
19. [D] Choose resistor values in Fig. 19 so that (a) X = 21 U3 + 13 U2 + 61 U1 and (b) the Thévenin
resistance between X and ground is 50 Ω. Why would the question be impossible if it had asked for
X = 12 U3 + 13 U2 + 13 U1 ?
Fig. 19
Fig. 20
20. [D] Choose resistor values in Fig. 20 so that (a) the Thévenin between X and ground is 50 Ω and
(b) it is possible to set X to 1, 2, 3 or 4 V by setting the switches appropriately.
21. [D] State whether the feedback in the following circuits is positive or negative:
(a)
Problem Sheet 3
(b)
Page 4 of 5
Ver 2379
E1.1 Analysis of Circuits (2014)
22. [D] Find an expression for Z in Fig. 22 in terms of U1 and U2 .
Fig. 22
Fig. 23
23. [D] The circuit of Fig. 23 is called a Howland current source. Show that the circuit has negative
feedback. Use nodal analysis to determine the current I and show that it does not depend on R.
√
24. [D] A non-linear device has a characteristic Y = X for inputs in the range 0 ≤ X ≤ 1. To improve
its linearity, the device is placed in a feedback loop using an op-amp with a gain of A as shown in
Fig. 24. Determine an expression for Y in terms
of U and A. Simplify the expression by using the
√
√ w
w2
− 8v
valid for v w. Estimate how large A
Taylor series approximation: v + w ≈ v 1 + 2v
2
must be to ensure that Y |U =0.5 = 0.5 × Y |U =1 to within 1% of Y |U =1 .
Fig. 24
25. [D] The diodes in Fig. 25 have a characteristic I = k exp VVT where VT = 25 mV.
(a) For the circuit of in Fig. 25(a), find an expression for U in terms of X assuming that X > 0.
(b) For the circuit of in Fig. 25(b), find and expression for Y in terms of X assuming that X > 0.
Fig. 25(a)
Fig. 25(b)
26. [C] We normally assume that the current at the inputs to an opamp are negligible. However this is not
always true and this question investigates the effect of non-zero input bias currents. If IB = 100 nA,
Y
find expressions for X
U and U in the circuits below. Explain the advantage of circuit (b) and derive
a general principal that should be followed when using opamps having non-negligible input bias
currents, IB .
(a)
Problem Sheet 3
(b)
Page 5 of 5
Ver 9024
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 3 - Solutions
1. (a) Thévenin voltage equals the open circuit voltage is 4 V (from potential divider). To obtain the
Thévenin/Norton we set the voltage source to 0 (making it a short circuit) and find the resistance of
4
= 5 A. This may also be found
the network to be 1||4 = 0.8 Ω. From this the Norton current is 0.8
directly from the short-circuit current of the original circuit.
(b) The open-circuit voltage is −8 V (since the 2 A current flows anticlockwise). To obtain the
Thévenin/Norton resistance, we set the current source to zero (zero current implies an open circuit),
so the resultant network has a resistance of 4 Ω. The Norton current is −8
4 = −2 A; this may also be
found by observing that the short-circuit current (flowing into node A) is +2 A.
A
2. KCL at node A gives A−5
1 + 4 − I = 0 from which 5A − 20 − 4I = 0 which we can rearrange to give
1
A = 4 + 0.8I = VT h + RT h I. We can also rearrange to give I = −5 + 0.8
A = −IN or + RN1or A.
3. (a) When U = 5, the diode is on and so X = U − 0.7 = 4.3. To check our assumption about the
diode operating region, we need to calculate the current through the diode; this equals 4.3 mA which
is indeed positive.
(b) When U = −5, the diode is off, so the current through the resistor is zero and X = 0. As a
check, the voltage across the diode is U − X = −5 which confirms our assumption that it is off.
4. As in part (a) of the previous question, I = 4.3 mA. The power dissipation in the diode is therefore
VD I = 0.7 × 4.3 = 3.01 mW. The power dissipation in the resistor is I 2 R = 18.5 mW.
10
5. Circuit (a) is an inverting amplifier with gain X
U = − 1 = −10. Circuit (b) is a non-inverting
amplifier with gain YU = 1 + 10
1 = +11. Another way to see this is to notice that, since the opamp
Y
inputs draw no current, the potential divider means that the -ve opamp input is at 11
and, since the
Y
negative feedback ensures the opamp terminals are at the same voltage, U = 11 .
6. [Method 1 - circuit manipulation] To calculate the Thévenin equivalent, we want to determine the
open-circuit voltage and the Thévenin resistance. To determine the open-circuit voltage, we assume
that I = 0 and calculate VAB . Since I = 0, we can combine the 1 Ω and 6 Ω resistors to give 7 Ω
and then combine this with the 6 Ω resistor in parallel to give 42
13 Ω. We now have a potential divider
42/13
98
so the voltage at point X is 63 × 3+42/13 = 3 . This is then divided by the 1 Ω and 6 Ω resistors
6
to give an open-circuit voltage of 98
3 × 7 = 28 V. The Thévenin/Norton resistance can be found by
short-circuiting the voltage source to give 3 Ω in parallel with 6 Ω which equals 2 Ω. This is then in
series with 1 Ω(to give 3 Ω ) and finally in parallel with 6 Ω to give 2 Ω.
[Method 2 - Nodal Analysis]. We can do KCL at node X (see diagram below) to get X−63
+X
3
6 +
X−A
= 0 which simplifies to 9X − 6A = 126 or 3X − 2A = 42. We now do KCL at A but include
1
an additional input current I as shown in the diagram. This gives A−X
+ A6 − I = 0 from which
1
7A − 6X = 6I. Substituting for 6X = 4A + 84 gives 3A = 84 + 6I or A = 28 + 2I. This gives
the Thévenin voltage as 28 and the Thévenin/Norton resistance as 2 Ω.Hence the Norton current is
14 A.
7. (a) KCL @ X gives
X−14
1
+
X
4
+
X
2
= 0 from which 7X = 56 ⇒ X = 8 ⇒ I =
X
2
= 4 mA.
(b)Finding the Thévenin equivalent of the left three components: we consider the two resistors as
a potential divider to give VT h = 14 × 54 = 11.2 V. Setting the source to zero (short circuit) gives
Th
RT h = 1||4 = 800 Ω. Hence I = RT hV+2000
= 11.2
2.8 = 4 mA.
Solution Sheet 3
Page 1 of 5
Ver 9024
E1.1 Analysis of Circuits (2016)
8. From question 7, the left three components have a Thévenin equivalent: VT h = 11.2 V and RT h =
800 Ω. It follows that the maximum power will be dissipated in R when R = RT h = 800 Ω (see notes
page 5-8). Since the voltage across R will then be 12 VT h the power dissipation will be 4R1T h VT2h =
39.2 mW.
9. (a) As shown in the sequence below, we first combine the current source with the 40 Ωresistor, then
combine the four series components into a single Thévenin equivalent and finally find the Thévenin
equivalent of the simple network (using parallel resistors for RT h and a potential divider for VT h ).
(b) Setting the voltage source to zero gives us the first diagram. Combining 40||(60 + 100) = 32 so
100
X = −0.1 × 32 = −3.2 V. it follows (potential divider) that A = −3.2 × 160
= −2 V.
Now setting the current source to zero gives the second diagram and we have a potential divider
100
= 3 V.
giving VAB = 6 × 100+60+40
Superposition now gives us VAB = VT h = −2 + 3 = 1 V.
To find RT h we set both sources to zero and find the resultant resistance of 100||(60+40) = 100||100 =
50 Ω.
10. (a) Negative, (b) Positive, (c) Negative, (d) Positive. In simple circuits like these, you can just see
which terminal the output feeds back to.
11. The best way to think of this circuit is as a potential divider with gain YU = 13 followed by a
60
X
1
7
non-inverting opamp circuit with gain X
Y = 1 + 10 = 7. The combined gain is then U = 3 × 7 = 3 .
Solution Sheet 3
Page 2 of 5
Ver 9024
E1.1 Analysis of Circuits (2016)
12. (a) You can either recognise
this a a standard inverting summing amplifier with gain
40
U1 + 40
U
X = − 20
=
−2U
2
1 − 4U2 or else apply KCL at the +ve input terminal with the assump10
tion that negative feedback will ensure that this terminal is at the same voltage as the –ve terminal
0−U2
0−X
1
i.e. 0 V. This gives: 0−U
20 + 10 + 40 = 0 from which X = −2U1 − 4U2 .
(b) The network connected to the +ve terminal is a weighted averaging circuit (page 3-7 of the notes)
50
so V+ = 31 U1 + 23 U2 . The opamp circuit itself is a non-inverting amplifier with a gain of 1 + 10
= 6.
2
1
So, Y = 6 × 3 U1 + 3 U2 = 2U1 + 4U2 .
(c) [Superposition
method] Following the method of part (b) above, if U3 = 0, we have Z = 5 ×
4
1
U
+
U
=
U
+ 4U2 . If, on the other hand, U1 = U2 = 0, then V+ = 0 and so we have an
1
2
1
5
5
inverting amplifier with a gain of − 40
10 = −4. Hence Z = −4U3 .
Combining these gives Z = U1 + 4U2 − 4U3 .
[Nodal analysis method] The top two resistors are a weighted average circuit so V+ = 51 U1 + 54 U2 .
Now, assuming that V− = V+ , we do KCL at V− to give
U1 + 4U2 − 4U3 − Z = 0 giving Z = U1 + 4U2 − 4U3 .
1
4
5 U1 + 5 U2 −U3
10
+
1
4
5 U1 + 5 U2 −Z
40
= 0 from which
13. We can use superposition. If U2 = 0,then V+ = 0 and we have an inverting amplifier with a gain
60
X
U1 = − R1 . The question tells us that this must equal −3 so we must have R1 = 20. Now, if U1 = 0,
the circuit consists of a potential divider with a gain of R260
+60 followed by a non-inverting amplifier
60
with a gain of 1 + R1 = 4. The combined gain must equal 2 (from the question) so the potential
divider must have a gain of 21 which means R2 = 60 kΩ.
14. The first opamp is non-inverting with a gain UX2 = 1 + 10
50 = 1.2. We can use superposition to find
Y : If U2 = 0,then X = 0 and we have a non-inverting amplifier with a gain of UY1 = 1 + 50
10 = 6.
Y
50
If, on the other hand, U1 = 0, then we have an inverting amplifier and X
= − 10
= −5. It
Y
follows that UY2 = X
× UX2 = −5 × 1.2 = −6. Combining both portions of the superposition,
Y = 6U1 − 6U2 = 6(U1 − U2 ). This is therefore a differential amplifier (whose output is ∝ (U1 − U2 ))
that draws almost no current from either if its inputs. A better circuit is given in question 22.
15. The potentiometer is a potential divider and so V+ = aU . Assuming that V− = V+ , we can do KCL
−X
at V− to get aU40−U + aU40
= 0 from which X = (2a − 1) U. For a = 0, the gain is −1 and when
a = 1, the gain is +1. So the circuit can generate any gain between these two extremes.
16. The Thévenin voltage is −3 V (potential divider) and the Thévenin resistance is 1||3 = 0.75 Ω as
shown in the diagram below. Note that if you drawn the Thévenin voltage source the other way
around (with “+” at the bottom) then the Thévenin voltage will be +3 V; this is an equally valid
solution.
X−(0−0.7)
(a) If U = 0 V, the diode is forward biassed and KCL @ X gives X−(−3)
= 0 from which
0.75 +
3
X = −2.54. The current through the diode is X−(−3)
=
613
mA
which
is
>
0
confirming
our guess
0.75
about the diode operating region.
(b) If U = 5 V, the diode is off and so X = −3 (since no current flows through the 750 Ω resistor).
The diode forward voltage is (−U ) − X = −2. This is < 0.7 confirming our operating region guess.
The diode switches regions when both operating region equations are true: ID = 0 and VD = 0.7.
The current equation implies X = −3 while the voltage equation (and the zero current through the
3 k resistor) implies X = −U − 0.7. Combining these gives U = 2.3. Extreme care with signs is
needed in this question.
Solution Sheet 3
Page 3 of 5
Ver 9024
E1.1 Analysis of Circuits (2016)
17. There is only one feedback loop in this circuit: from W back to the input adder. We can write
FG
W = F G (X − W ) from which W = 1+F
G X. Then Y = F (X − W ) + HW . From the first equation
1
FG
F +F GH
W
(X − W ) = F G so we can substitute this in to get Y = W
G + HW = G + H 1+F G X = 1+F G X.
−A
18. (a) [Nodal analysis including I ] We have J = U100
, so the current source value (expressed in terms
of node voltages) is 49J = 0.49 (U − A). The KCL equation @ A is A−U
100 − 0.49(U − A) − I = 0 from
which 50 (A − U ) = 100I which gives A = 2I + U = RT h I + VT h . So the Thévenin voltage is U and
the Thévenin resistance is 2 kΩ.
(b) In this method, we assume I = 0 and calculate the open-circuit voltage and short-circuit current.
For the open-circuit voltage, we do KCL at A and obtain A−U
100 − 0.49(U − A) = 0 from which
50 (A − U ) = 0 so A = VT h = U . For the short-circuit current, we join A and B and the current is
VT h
U
then R
= 100
+ 0.49U = 0.5U . Hence RT h = 2 kΩ.
Th
2 G2 +U3 G3
. The equations are made much easier to solve
19. From the notes (page 3-7) X = U1 GG1 +U
1 +G2 +G3
because we know that the Thévenin resistance must be 50 Ω. The Thévenin resistance is just
1
. Hence X = 21 U3 + 13 U2 + 16 U1 =
the parallel combination R1 ||R2 ||R3 = G1 +G12 +G3 = 20 mS
50 (U1 G1 + U2 G2 + U3 G3 ) where the first expression is given in the question and the second comes
1
from substituting for G1 + G2 + G3 . Identifying the coefficients in this equation gives G1 = 300
,
1
1
G2 = 150 and G3 = 100 from which R1 = 300, R2 = 150 and R3 = 100. In a weighted average
circuit, the coefficients must sum to 1; thus 12 + 31 + 61 = 1 but 12 + 13 + 13 6= 1 so the latter set of
coefficients is inadmissible.
2 +U3 G3 +5G4
. where U2 and U3 can equal 0 or 5. As in the previous
20. From the notes (page 3-7) X = UG2 G
1 +G2 +G3 +G4
question, we are told that G1 + G2 + G3 + G4 = 20 mS. Hence X = 50 (U2 G2 + U3 G3 + 5G4 ). The
only way that we can obtain the required voltages is if switching U2 causes X to change by 1 V and
switching U3 causes X to change by 2 V (or vice versa). Thus we need X = 51 U2 + 25 U3 + 1; it is easy
to see that this satisfies the required output voltages. Now equating coefficients, we get 50G2 = 51 ,
50G3 = 25 and 250G4 = 1 from which R2 = G12 = 250, R3 = G13 = 125 and R4 = G14 = 250. Finally,
G1 = 0.02 − G2 − G3 − G4 = 0.004 so R1 = G11 = 250.
21. We need to determine how the differential input to the opamp (V+ − V− )depends on X. We are not
interested in how it depends on U so it is convenient to set U = 0. So, with this assumption, we get
potential divider equations:
20
10
(a) V+ = 10+20
X = 23 X and V− = 20+10
X = 31 X. Hence V+ − V− = 31 X which, since the coefficient
is positive means positive feedback.
10
20
(b) V+ = 10+20
X = 13 X and V− = 20+10
X =
coefficient is negative means negative feedback.
2
3 X.
Hence V+ − V− = − 31 X which, since the
22. We can split the circuit up into two independent parts because the opamp outputs are voltage sources
whose voltage is not affected by how many other things are connected to them. Note that all three
opamps have negative feedback and so we can assume that V+ = V− .
−X
−U2
For the first part, we have A = U1 and B = U2 . KCL @ A therefore gives U150
+ U110
= 0 which
U2 −Y
U2 −U1
gives X = 6U1 − 5U2 . Similarly, KCL @ B gives 50 + 10 = 0 which gives Y = 6U2 − 5U1 .
The second part of the circuit is a differential amplifier (see page 6-9 of the notes) for which Z =
60
20 (Y − X). Substituting in the expressions for X and Y gives Z = 3 ((6U2 − 5U1 ) − (6U1 − 5U2 )) =
3 (11U2 − 11U1 ) = 33 (U2 − U1 ).
Solution Sheet 3
Page 4 of 5
Ver 9024
E1.1 Analysis of Circuits (2016)
23. We can verify that the opamp has negative feedback since (if we set the 10 V source to zero), we
2||R
have V− = 21 Y but V+ = αY where α = 2+2||R
< 12 . So, we can assume that V+ = V− = X. KCL @
X
X−Y
+ X−Y
+ I = 0.
V− gives X−(−10)
10
10 = 0 which gives X − Y = −X − 10. KCL @ V+ gives 2 +
2
−X−10
X
Substituting X − Y = −X − 10 gives 2 + 2 + I = 0 which simplifies to I = 5. Notice that this
does not depend on R, so we have constructed a current source.
√
√
2
24. From the block diagram, we can deduce Y = X = AU
√ − AY from
which Y + AY − AU = 0.
2
Solving this quadratic equation gives Y (U ) = 0.5 −A + A + 4AU . Notice that only one of the
two roots will result in Y being positive.
Provided
that A 4U , we can use the Taylor series
2
2U 2
approximation to give Y (U ) ≈ 0.5 −A + A 1 + 2U
−
= U − UA . From this, Y (1) = 1 − A1
A
A2
0.25
and Y (0.5) = 0.5 − 0.25
A = 0.5Y (1) + A . We would like the error,
0.25
1
A ≤ 0.01 1 − A . Solving this gives A ≥ 26.
0.25
A
to be 1% of Y (1), i.e.
25. (a) The − terminal of the opamp is a virtual earth so the current through the resistor is X
R . All the
−U
from
which
current flows through the diode whose voltage is −U. Therefore we have X
=
k
exp
R
VT
X
U = −VT ln kR
.
X
(b) The second opamp is a non-inverting amplifier with a gain of 3 so W = 3U = −3VT ln kR
. For
Y
−W
X
X 3
the third opamp, the current thorough the diode is 2R = k exp VT = k exp 3 ln kR = k kR
.
2
3
From this we find that Y = k2 R2 X . This we have made a circuit that cubes its input voltage (times
a scale factor).
26. (a) Despite the current IB , it is still the case that V+ = 0 and so, because of the negative feedback,
0−Y
P = 0 also. KCL @ P gives 0−U
10 +IB + 40 = 0 from which −4U +40IB −Y = 0 or Y = −4U +40IB .
Substituting IB = 0.0001 mA gives Y = −4U + 0.004 so there is an output error of 4 mV.
(b) This time, the current IB flows through the 8 kΩ resistor so V+ = −8IB . As before, we assume
Q−Y
that Q = V+ also. Then KCL @ Q gives Q−U
10 + IB + 40 = 0 from which −4U + 40IB − Y + 5Q = 0.
Substituting Q = −8IB gives −4U + 40IB − Y − 40IB = 0 which simplifies to Y = −4U . Thus in
this circuit, the bias currents do not cause any error.
The moral is that when designing opamp circuits, you should try to make the Thévenin resistance
seen by the two input terminals the same. If you achieve this, and if the bias currents are the same
at both inputs (usually approximately true) there will be no resultant errors.
Solution Sheet 3
Page 5 of 5
Ver 5542
E1.1 Analysis of Circuits (2015)
E1.1 Circuit Analysis
Problem Sheet 4 (lectures 9 & 10)
Key: [A]= easy ... [E]=hard
Note: A “dimensioned sketch” should show the values on the x and y axes corresonding to significant
places on the corresponding graph.
1. [B] For each of the following waveforms, determine the corresponding phasor in both the form a + jb
and r∠θ.
(a) 8 cos ωt.
(b) 3 cos ωt + 4 sin ωt.
(c) 2 cos ωt + π4 .
(d) 8 sin ωt.
(e) −2 cos ωt.
(f) −4 sin ωt −
π
2
π
4
+ 5 sin ωt −
(g) 8 cos ωt +
.
π
4
.
2. [B] For each of the following phasors, determine the corresponding waveform in both the form
π
a cos ωt + b sin ωt and a cos (ωt + θ). (a) 1, (b) −2, (c) 3j, (d) −4j, (e) j − 1, (f) 3 − 4j, (g) 2ej 2 ,
π
(h) 4e−j 6 .
3. [B] For each of the following cases say which of the two waveforms or phasors leads the other:
(a) sin ωt and cos ωt.
(b) sin (ωt + π) and cos ωt.
(c) sin (ωt − π) and cos ωt.
(d) (1 + j) and (2 + j).
(e) (1 + j) and (1 − j).
(f) (−1 + j) and (−1 − j).
(g) 1 and 1∠350◦ .
4. [B] Draw a dimensioned sketch of the waveform of i in the circuit of Fig. 4(a) when v has the
waveform shown in Fig. 4(b).
Fig. 4(a)
Fig. 4(b)
5. [B] For each of the circuits shown in Fig. 5(a)-(d) determine the average value of y(t) when x(t) =
4 + 2 cos ωt for some non-zero frequency ω.
Fig. 5(a)
Fig. 5(b)
Fig. 5(c)
Fig. 5(d)
6. [B] Find the value of a single inductor equivalent to the circuit shown in Fig. 6.
Problem Sheet 4
Page 1 of 3
Ver 5542
E1.1 Analysis of Circuits (2015)
Fig. 6
Fig. 7
7. [B] Find the value of a single capacitor equivalent to the circuit shown in Fig. 7 given that each of
the capacitors has a value of 1 µF.
8. [B] Find the average value of v in the circuit of Fig. 8 if u(t) = 2 + 3 cos ωt.
Fig. 8
Fig. 9
9. [B] Find the average value of v in the circuit of Fig. 9 if u(t) = 8 − 2 cos ωt.
10. [B] Find the complex impedance of the circuit shown in Fig. 10 for (a) ω = 0, (b) ω = 1000, (c)
ω = 2000 and (d) ω = ∞.
Fig. 10
Fig. 11(a)
Fig. 11(b)
Fig. 11(c)
11. [B] The components in Fig. 11 are labelled with their impedances. Calculate both the complex
impedance and the complex admittance for each of the three networks.
12. [B] The components in Fig. 12 are labelled with their impedances. Determine the values of a parallel
inductor and resistor that will have the same overall impedance at (a) 1 kHz and (b) 10 kHz. Hint:
first calculate the admittance of the original network.
Fig. 12
Fig. 13(a)
Fig. 13(b)
13. [C] Draw a dimensioned sketch of the waveform of v(t) in the circuit of Fig. 13(a) when i has the
waveform shown in Fig. 13(b).
Problem Sheet 4
Page 2 of 3
Ver 5542
E1.1 Analysis of Circuits (2015)
√
π
14. [C] The three current i1 , i2 , i3 in Fig. 14 are equal to 5 cos ωt + 3π
8 cos ωt
4 , 2 cos ωt + 4 and
but not necessarily in that order. Determine which current is which and find both the phasor I and
time-waveform i(t).
Fig. 14
Fig. 17(a)
Fig. 17(b)
15. [D] Draw a dimensioned sketch of the waveform of i in the circuit of Fig. 15 when v has the waveform
shown in Fig. 4(b) given that at time t = 0, (a) i(0) = 0 and (b) i(0) = 2 A.
16. [D] Draw a dimensioned sketch of the waveform of v in the circuit of Fig. 16(a) when i has the
waveform shown in Fig. 16(b) given that at time t = 0, (a) v(0) = 0 and (b) v(0) = −5 V.
Fig. 15
Fig. 16(a)
Fig. 16(b)
Fig. 18
17. [D] In the circuit of Fig. 17(a), the voltage v has the periodic waveform shown in Fig. 17(b) with a
period of 4 µs and an amplitude of 20 V.
(a) State the duty cycle of v.
(b) Determine the average value of x.
(c) Determine the average value of iR .
(d) Determine the average value of iL .
(e) Assuming that x is constant (at its average value), draw a dimensioned sketch of the waveform
of iL (t) and determine its maximum and minimum values.
(f) Assuming that x is constant (at its average value), determine the average, positive peak and
negative peak of the powers absorbed by each of R, L and C.
18. [D] In the circuit of Fig. 18, the output logic levels from the inverter are 0 V and 5 V and the inverter
has a maximum output current of ±2 mA. The inverter senses a low voltage when its input is < 1.5 V.
If x changes from logic 0 to logic 1, determine the delay until z changes. Ignore the inverter input
currents and any delays inside the inverters themselves.
Problem Sheet 4
Page 3 of 3
Ver 6647
E1.1 Analysis of Circuits (2015)
E1.1 Circuit Analysis
Problem Sheet 4 - Solutions
1. (a) 8, (b) 3 − j4 = 5∠ − 0.93 (−53◦ ), (c) 1.4 + j1.4 = 2∠0.79 (45◦ ), (d) −j8 = 8∠ − 1.57 (90◦ ),
(e) −2 = 2∠3.14 (180◦ ), (f) 4, (g) 2.12 + j2.12 = 3∠0.79 (45◦ ).
2. (a) cos ωt, (b) −2 cos ωt = 2 cos (ωt + π), (c) −3 sin ωt = 3 cos ωt + π2 , (d) 4 sin ωt = 4 cos ωt − π2 ,
3π
(e) − cos ωt −
sin ωt = 1.4 cos ωt + 4 , (f) 3 cos ωt +
4 sin ωt = 5 cos (ωt − 0.93), (g) −2 sin ωt =
π
2 cos ωt + 2 , (h) 3.46 cos ωt + 2 sin ωt = 4 cos ωt − π6 .
3. (a) cos ωt by π2 , (b) sin (ωt + π) by π2 , (c) sin (ωt − π) by π2 : note that sin (ωt + π) and sin (ωt − π) are
actually the same waveform, (d) (1 + j) by 0.322 rad, (e) (1 + j) by π2 , (f) (−1 − j) by π2 , (g) 1 by 10◦ .
Because angles are only defined to within a multiple of 360◦ , you always need to be careful when
comparing them. To find out which is leading, you need to take the difference in phase angles and
then add or subtract multiples of 360◦ to put the answer into the range ±180◦ . Note that a sine
wave is defined for all values of t (not just for t > 0) and so there is no such thing as the “first peak”
of a sine wave.
dv
4. i = C dv
dt . dt is 3000 V/s for the first 4 ms and −6000 V/s for the next 2 ms. So i = +15 or −30 mA.
(see Fig. 4)
20
C
0
-20
-40
0
2
4
6
t (ms)
Fig. 4
5. The average value of x(t) is X = 4 (note that we use capital letters for quantities that do not
vary with time). For averages (or equivalently for DC or ω = 0) capacitors act as open circuit and
inductors as short circuits; this gives the simplified circuits shown below. So this gives (a) Y = 3,
(b) Y = X = 4, (c) Y = X = 4, (d) Y = 12 X = 2.
Fig. 5(a)
Fig. 5(b)
Fig. 5(c)
Fig. 5(d)
6. 4 + 4 = 8. 8||8 = 4. 4 + 4 = 8. 24||8 = 6 mH.
7. C9 and C10 are short-circuted and play no part in the circuit. We can merge series and parallel
capacitors as follows: C4,5 = 0.5, C7,8 = 2, C2,3 = 2. Now merge C4,5 with C6 to give C4,5,6 =
1.5 and merge this with C7,8 = 2 to give C4,5,6,7,8 = 67 . Now merge this with C2,3 = 2 to give
20
C2,3,4,5,6,7,8 = 20
7 . Finally merge this with C1 = 1 to give C = 27 µF.
8. To determine average values, we can treat C as open circuit and L as short circuit. The original
circuit simplifies to that shown in Fig. 8. So we have a simple potential divider and v̄ = 12 ū = 1 V
where the overbar denotes “average value”.
Fig. 8
Solution Sheet 4
Fig. 9
Page 1 of 3
Ver 6647
E1.1 Analysis of Circuits (2015)
9. To determine average values, we can treat C as open circuit and L as short circuit. The original
circuit simplifies to that shown in Fig. 9 and so v̄ = ū = 8 V where the overbar denotes “average
value”.
10. Z = RS + RP ||jωL. (a) Z = RS = 10, (b) Z = 10 + 10000||100j = 11 + 100j = 100.6∠83.7◦ ,
(c) Z = 10 + 10000||200j = 14 + 200j = 200.4∠86◦ , (d) Z = 10010.
11. We denote the impedance by Z and the admittance by Y = Z1 . (a) Z = 1.44 + 1.92j and Y =
0.25 − 0.33j, (b) Z = 4 − 3j and Y = 0.16 + 0.12j, (c) Z = 4 and Y = 0.25. Notice that the
imaginary part of the impedance (the reactance) is positive for inductive circuits and negative for
capacitive circuits, but that the imaginary part of the admittance (the susceptance) has the opposite
sign. In part (c), the impedances of the inductor and the capacitor have cancelled out leaving
an overall impedance that is purely real; because the impedances are frequency dependent, this
cancellation will only happen at one particular frequency which is called the network’s “resonant
frequency”. I strongly advise you to learn how to do these complex arithmetic manipulations using
the built-in capabilities of the Casio fx-991.
12. (a) ω = 6283 so the impedance is Z = 100 + 314j. Taking the reciprocal gives Y = Z1 = 0.92 −
2.89j mS. Since parallel admittances add, the parallel component values must be RP = 1000
0.92 = 1087Ω
1000
= 55.1 mH. For case (b), we now have ω = 62832 and, following the same argument,
and LP = 2.89ω
we get RP = 98.8 kΩ and LP = 50.05 mH. As the frequency goes up, the series resistor becomes a
less significant part of the total impedance and its effect becomes less. This means that LP becomes
approximately equal to the original inductance and RP becomes larger.
di
13. v = L dt
.
Fig. 13)
di
dt
is 3 A/s for the first 4 ms and −6 A/s for the next 2 ms. So v = +6 or −12 mV. (see
14. If we define the voltage phase to be φ will be ∠i1 = φ + 0, φ − π2 < ∠i2 < φ + 0, ∠i3 = φ + π2 as
the CIVIL mnemonic
√ us. Thus i3 will have the most
reminds
positive phase πshift. It follows that
i1 = 2 cos ωt + π4 , i2 = 8 cos ωt and i3 = 5 cos ωt + 3π
and that φ = 4 . As phasors these
4
are I1 = 1.4 + j1.4 = 2∠45◦ , I2 = 2.8, I3 = −3.5 + j3.5 = 5∠135◦ . Adding these together gives
I = 0.71 + j4.95 = 5∠1.43 (82◦ ). So i(t) = 0.71 cos ωt − 4.95 sin ωt = 5 cos (ωt + 1.43) A.
´t
15. i(t) = i(0) + L1 τ =0 v(τ )dτ where v(τ ) = 3000τ for 0 ≤ τ ≤ 4 ms and v(τ ) = 36 − 6000τ for
4 ms ≤ τ ≤ 6 ms (obtain this formula by finding the straight line equalling 12 at τ = 4 ms and 0 at
τ = 6 ms). So, for the first segment,
i = L1 × 1500t2 which reaches 12 A at t = 0.004. For the second
1
2
segment, i = L × 36t − 3000t + c. To find c, we force i = 12 at t = 0.004. This gives c = −36 A.
When t = 0.006 we then get i = 18 A. For part (b), we just add 2 A to the curve. (see Fig. 15)
´
16. v = C1 idt where i = 3t or i = 36 × 10−3 − 6t. So, for the first segment, v= C1 × 1.5t2 which reaches
4.8 V at t = 0.004. For the second segment, v = C1 × 36 × 10−3 t − 3t2 + a. To find a, we force
v = 4.8 at t = 0.004. This gives a = −14.4 V. When t = 0.006 we then get v(t) = 7.2 V. For part
(b), we just subtract 5 V from the curve. (see Fig. 16)
10
10
20
5
0
10
0
0
L
C
30
0
2
4
t (ms)
Fig. 15
6
-5
0
2
4
t (ms)
Fig. 16
6
-10
-20
0
2
4
6
t (ms)
Fig. 13
x̄
17. (a) The duty cycle is 0.25 = 25%, (b) x̄ = v̄ = 14 × 20 = 5 V, (c) īR = R
= 5 mA, (d) Since
diL
īC = 0, īL = īR = 5 mA, (e) The voltage across the inductor is v − x = L dt . So when v = 20,
diL
15
dt = L = 7.5 kA/s. So the total change in iL over the 1 µs interval is 7.5 mA. It follows that
iL varies from its average value of 5 mAby ±3.75 mAand has minimum and maximum values of 1.25
and 8.75 mA (see Fig. 17(a)). (f) Average powers are PR = 25 mW, PL = PC = 0. Max powers
are PR = 25 mW, PL = vL iL = 15 × 8.75 = 131.25 mW, PC = vC iC = 5 × 3.75 = 18.75 mW.
Min powers are PR = 25 mW, PL = vL iL = −5 × 8.75 = −43.75 mW (see Fig. 17(b)), PC =
vC iC = 5 × −3.75 = −18.75 mW (see Fig. 17(c)). Note that during the time that it is positive
Solution Sheet 4
Page 2 of 3
Ver 6647
E1.1 Analysis of Circuits (2015)
(0.5 < t < 2.5 ms), the average value of iC is ıC = 1.375 mA and so the total rise in vC will be
= 275 mV (i.e. ±138mVaround its mean) which is small compared to its
∆vC = ıCC∆t = 1.375×2
10
mean value of 5 V; this justifies the assumption that it is constant.
10
20
100
50
5
0
0
0
0
2
4
t (µs)
6
Fig. 17(a)
8
0
2
4
t (µs)
6
Fig. 17(b)
8
-20
0
2
4
t (µs)
6
8
Fig. 17(c)
18. When x changes from low to high, y will change from high to low. The maximum current is 2 mA
i
3.5
−6
so dy
= 70 ns.
dt = − C = −50 MV/s. So the time to fall from 5 V to 1.5 V is 50 × 10
Solution Sheet 4
Page 3 of 3
Ver 7844
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 5 (Lectures 11, 12 & 13)
Key: [A]= easy ... [E]=hard
Note: A “sketch” should show the values on the x and y axes corresonding to significant places on
the corresponding graph.
1. [C] For each of the circuits in Fig. 1(i)-(vi),
(a) Find the transfer function
Y
X
(jω).
(b) Find expressions for the low and high frequency asymptotes of H (jω).
(c) Sketch the straight line approximation to the magnitude response, |H (jω)|, indicating the
frequency (in rad/s) and the gain of the approximation (in dB) at each of the corner frequencies.
Fig. 1(i)
Fig. 1(ii)
Fig. 1(iii)
Fig. 1(iv)
Fig. 1(v)
Fig. 1(vi)
2. [C] Sketch a straight line approximation for the phase response, ∠H (jω), of the circuit in Fig. 1(v),
indicating the frequency (in rad/s) and phase (in rad) at each of the corner frequencies.
2
k(jω)
where a = 129
3. [B] A “C-weighting filter” in audio engineering has the form H (jω) = (jω+a)
2
(jω+b)2
and b = 76655 rad/s. Calculate k exactly so that |H (jω)| = 1 at ω = 2000π rad/s.
4. [C] Design circuits with each of the magnitude responses given in Fig. 4(i)-(iii) using, in each case,
a single capacitor and appropriate resistors.
Fig. 4(i)
Fig. 4(ii)
Fig. 4
5. [B] Express each of the following transfer functions in a standard form in which thenumerator and de
2
nominator are factorized into linear and quadratic factors of the form (jω+p) and (jω) + qjω + r2
where p, q and r are real with an additional numerator factor of the form A(jω)k if required. Quadratic terms should be factorized if possible.
(a)
−2ω 2 −2jω 3
1−2ω 2 +ω 4
(b)
−2(1+ω 2 )
(1−ω 2 )+2jω
(c)
10(jω)2 +2jω+10
(jω)2 +2jω+1
(d)
1
jω+6(jω)−1 +5
Problem Sheet 5
Page 1 of 3
Ver 7844
E1.1 Analysis of Circuits (2016)
6. [B] Without doing any algebra, determine the low and high frequency asymptotes of the following
transfer functions:
(a)
−2ω 2 −2jω 3
1+ω 4
(b)
2(jω)3 +3
4(jω)4 +1
(c)
(d)
jω (2(jω)6 +3)(5(jω)3 +4jω+3)
2((jω)5 +1)((jω)5 +5)
12
jω+6(jω)−1
α
7. [D] A circuit has a transfer function whose low and high frequency asymptotes are A (jω) and
β
B (jω) respectively. What constraints can you place on α and β if you know that the magnitude of
the transfer function is less than G at all frequencies where G is a fixed real-valued constant.
8. [C] For each of the following transfer functions, sketch the straight line approximation to the magjω/500)
nitude response and determine the gain of this approximation at ω = 1000 rad/s: (a) (1+jω5(1+
/100)(1+jω/2000) ,
(b)
2(1+jω/5000)
(1+jω/100) ,
(c)
3jω(1+jω/500)
(1+jω/100)(1+jω/2000)(1+jω/5000) .
2
( jω
p )
where the corner
( ) +2ζ ( jω
p )+1
√
1
frquency, p = ζRC rad/s. Using capacitors of value C = 10 nF, design a filter with ζ = 0.5 and a
p
ω
corner frequency of 2π
= 1 kHz. Determine the value of the transfer function at 2π
= 100 Hz, 1 kHz
and 10 kHz.
9. [C] The frequency response of the circuit in Fig. 9 is given by
jω
P
2
10
Gain (dB)
0
-10
-20
A
B
C
-30
-40
Fig. 9
Fig. 11(a)
1
2
5
10
20
Frequency (krad/s)
50
100
200
Fig. 11(b)
10. [D] A high-pass “Butterworth” filter of order 2N consists of N cascaded copies
of Fig. 9 all having
the same corner frequency, but with the k th stage having ζk = cos 2k−1
π where k = 1, . . . , N .
4N
(“cascaded” means you connect the output of one stage to the input of the next). Design a 4th order
high-pass Butterworth filter with a corner frequency of 1 kHz. Write an expression for its transfer
function and sketch its magnitude response using just the high and low asymptotes. Butterworth
filters are widely used because they have a very smooth magnitude response without any peaks; the
4N
(ω)
2
transfer function satisfies |H(jω)| = ω p 4N .
( p ) +1
11. [C] The circuit of Fig. 11(a) is a high-pass filter whose magnitude response is marked “A” in Fig. 11(b)
. Using the filter transformations described in lectures, design filters with the magnitude responses
marked “B” and “C” on the graph. Relative to “A”, these are respectively shifted up in frequency by
a factor of 5 and reflected in the axis ω = 10, 000.
12. [C] (a) Find the resonant frequency, ωr , at which the impedance of the network in Fig. 12(i) is real.
(b) Determine the Q of the circuit at ωr . (c) Find RP and LP in the circuit of Fig. 12(ii) so that
the two networks have the same impedance at ωr .
Fig. 12(i)
Problem Sheet 5
Fig. 12(ii)
Fig. 13
Page 2 of 3
Ver 7844
E1.1 Analysis of Circuits (2016)
13. [C] In the circuit of Fig. 13, ω = 10 000 and the phasor V = 10. Find (a) the peak power supplied
by V and (b) the peak power absorbed by C.
14. [C] Determine the transfer function for each of the circuits Fig. 14(i)-(iv).
Fig. 14(i)
Fig. 14(ii)
Fig. 14(iii)
Fig. 14(iv)
15. [D] For the circuit in Fig. 15,
(a) Find the transfer function
Z
W
(jω) and explain why this is equal to
Y
W
(jω)
(b) Hence, by applying KCL at node W and using part (a) to substitute for W , show that the
Y
1
transfer function X
(jω) = R R C C (jω)2 +(R
.
+R )C jω+1
1
2
1
2
1
2
1
(c) From the transfer function expression we can express the corner frequency and damping factor
2 )C1
as p2 = R1 R21C1 C2 and ζ = p(R1 +R
. By eliminating p between these equations, show that
2
q
2
1
1 + 1 − ζC
= 2. Explain why this means that we must choose C1 ≤ ζ 2 C2 .
1+ R
2C
R1
2
(d) Assuming that you only have available capacitors of values 10 nF, 22 nF and 47 nF, design a
2
filter with p = 1000 × 2π and ζ = 0.5. Choose C1 and C2 first. then R
R1 and lastly R1 .
16. [D] For the circuit of Fig. 16,
(a) Find the transfer function
Y
X
(jω).
Y
(b) Find the frequency, ω0 , at which X
(jω) is maximum and its value at this maximum.
(c) Find the 3dB bandwidth of the circuit and the value of Q =
1
2ζ .
[Note: If you find yourself doing loads of algebra, you are using the wrong method.]
Fig. 15
Problem Sheet 5
Fig. 16
Page 3 of 3
Ver 9120
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 5 - Solutions
1. Each of the circuits may be viewed as a potential divider, so we can write down the transfer function
without doing any nodal analysis. In two cases, one element of the potential divider consists of a parR
1
= 1+jωRC
. Graphs of
allel R||C combination. This parallel combination has the impedance 1/R+jωC
the magnitude responses are shown in Fig. 1(i)-(vi) together with their straight-line approximations.
/500
jωRC
Y
(i) X
= R+1R/jωC = 1+jωRC
= 1+jω
/500 where RC = 2 ms. LF asymptote is 0.002jω; HF asymptote
1
is 1. Denominator corner frequency is RC
= 500 rad/s. At the corner, the gain is 1 = 0 dB.
jω
(ii)
Y
X
1/jωC
R+1/jωC
−1
=
500 (jω)
=
1
1+jωRC
=
1
1+jω/500
where RC = 2 ms. LF asymptote is 1; HF asymptote is
1
RC
. Denominator corner frequency is
= 500 rad/s. At the corner, the gain is 1 = 0 dB.
jω L/R
1+jω L/R
jωL
R+jωL
Y
L
(iii) X
=
=
where R
= 100 µs. LF asymptote is 10−4 jω; HF asymptote is 1.
R
Denominator corner frequency is L = 104 rad/s. At the corner, the gain is 1 = 0 dB.
Y
X
(iv) For convenience, we define R = 1 k. Then
=
R+jωL
5R+jωL
L
1+jω R
= 0.2 1+jω
R
L
where
L
5R
L
R
= 1 µs. LF
asymptote is 0.2; HF asymptote is 1. Numerator corner frequency is
= 10 krad/s with a gain at
5R
the corner of 0.2 = −14 dB. Denominator corner frequency is L = 50 krad/s with a gain at the
corner of 1 = 0 dB. As can be seen in Fig. 1(iv), the magnitude response turns up at 10 krad/s and
then flattens out again at 50 krad/s . In between these two frequencies the slope log|H|
log ω = +1 or,
equivalently, +6 dB/octave or +20 dB/decade ; all these are the same as saying that |H| ∝ ω; thus
from ω = 10 krad/s to 50 krad/s, the frequency increases by a factor of 5 and the gain also increases
by a factor of 5.
Y
X
(v) For convenience, we define R = 1 k. Then
=
R
8R
2R+ 1+8jωRC
=
1+8jωRC
10+16jωRC
1+8jωRC
= 0.1 1+1.6jωRC
where RC = 100 µs. LF asymptote is 0.1; HF asymptote is 0.5. Numerator corner frequency is
1
8RC = 1250 rad/s with a gain at the corner of 0.1 = −20 dB. Denominator corner frequency is
1
1.6RC = 6.25 krad/s with a gain at the corner of 0.5 = −6 dB. As in the previous part, both the
frequency and the gain change by a factor of 5 between the corner frequencies.
(vi) For convenience, we define R = 10 k. Then
Y
X
=
R
1+jωRC
1
R
2R+ jωC
+ 1+jωRC
jωRC
1+4jωRC+2(jωRC)2
=
jωRC
2jωRC(1+jωRC)+(1+jωRC)+jωRC
−1
0
0
-5
-5
-10
-15
-20
-10
-15
-20
3
ω (rad/s)
-5
-10
|H| (dB)
-5
-10
-15
-20
ω (rad/s)
3
10
4
10
5
10
ω (rad/s)
10
Fig. 1(iii)
-10
-15
-20
-15
-25
6
10
7
ω (rad/s)
Fig. 1(iv)
Solution Sheet 5
-15
Fig. 1(ii)
0
-25
5
10
3
10
Fig. 1(i)
-10
-20
2
10
|H| (dB)
2
10
|H| (dB)
|H| (dB)
0
-5
|H| (dB)
|H| (dB)
where RC = 1 ms. LF asymptote is 0.001jω; HF asymptote is 500 (jω) . We
2
can factorize the denominator to give 1 + 4jωRC + 2 (jωRC) = 1 + jω
1 + jω
where a and
a
b
√
b are −1 times the roots of the quadratic equation 2R2 C 2 x2 + 4RCx + 1 or 1±RC0.5 . This gives
denominator corner frequencies a = 293 and b = 1707 rad/s. The gain in between these two frequencies can√be obtained by substituting ω = a into the LF asymptote expression to give a value of
aRC = 1 − 0.5 = 0.293 = −10.7 dB.
10
-30
3
10
4
ω (rad/s)
Fig. 1(v)
10
-20
2
10
3
10
ω (rad/s)
4
10
Fig. 1(vi)
Page 1 of 7
=
Ver 9120
E1.1 Analysis of Circuits (2016)
2. For convenience, we define R = 1 k. Then
Y
X
=
R
8R
2R+ 1+8jωRC
1+8jωRC
10+16jωRC
=
1+8jωRC
= 0.1 1+1.6jωRC
where
0.8
0
0.6
-10
|H| (dB)
arg(H) (rad)
RC = 100 µs. LF asymptote is 0.1; HF asymptote is 0.5; both of these are real and so have
1
zero phase shift. The magnitude plot has a numerator corner frequency of 8RC
= 1250 rad/s
1
and denominator corner frequency of 1.6RC = 6.25 krad/s . Each of these generates a pair of
corner frequencies on the phase plot at 0.1× and 10× the frequency. Thus we have corners at
ω = 125 (+), 625 (−), 12.5 k (−), 62.5 k (+) rad/s where the sign in parentheses indicates the gradient change ± π4 rad/decade. Between 125 and 625 rad/s the gradient is π4 rad/decade so the phase
π
will change by π4 × log10 625
125 = 4 × 0.7 = +0.55 rad. This is therefore the phase shift for the flat part
of the phase response. (see Fig. 2).
0.4
0.2
-20
-30
0
-40
3
10
2
4
3
10
10
ω (rad/s)
Fig. 2
10
5
10
Fig. 3
3. At ω = 2000π = 6283.2 we know |H(jω)| = 1. Hence k =
3.9495×107 ×5.9155×109
3.9478×107
4
10
ω (rad/s)
|jω+a|2 |jω+b|2
|jω|2
=
(ω2 +a2 )(ω2 +b2 )
ω2
=
9
= 5.918 × 10 .
4. (i) This is the same low-pass filter as Fig. 1(ii) but with a corner frequency of 50 rad/s. So we want
1
RC = 50
= 20 ms. One possible choice is shown in Fig. 4(i).
(ii) This is the same high-pass filter as Fig. 1(i) but with a corner frequency of 1000 rad/s and a high
1
frequency gain of 0.5 = −6 dB. So we want RC = 1000
= 1 ms. One possible choice is shown in Fig.
4(ii); the two resistors give the correct high frequency gain.
(iii) We want a circuit whose gain decreases from 12 at low frequencies to 18 at high frequencies. We
can do this by using a capacitor to short out part of the vertical limb of the potential divider at
high frequencies as shown in Fig. 4(iii). This design has a gain of 18 when the capacitor is a short
circuit; with the capacitor open circuit (low frequencies), we add in an additional 6R which gives a
gain of
1
2.
The impedance of 6R||C is
6R
R+ 1+6jωRC
6R
8R+ 1+6jωRC
=
7+6jωRC
14+48jωRC
which, as a check, we see has the
7
correct LF and HF asymptotes. The numerator corner frequency is at ω = 6RC
which needs to be
at 1000 rad/s. From this, RC = 1.17 ms so one possible set of value is C = 100 nF and R = 12 kΩ.
Fig. 4(i)
5. (a)
−2ω 2 −2jω 3
1−2ω 2 +ω 4
=
2(jω)2 (jω+1)
((jω)2 +1)((jω)2 +1)
(b)
−2(1+ω 2 )
(1−ω 2 )+2jω
(c)
10(jω)2 +2jω+10
(jω)2 +2jω+1
(d)
1
jω+6(jω)−1 +5
Solution Sheet 5
Fig. 4(ii)
2(jω+1)(jω−1)
(jω+1)(jω+1)
=
=
=
=
Fig. 4
.
2(jω−1)
(jω+1)
10((jω)2 +0.2jω+1)
(jω+1)(jω+1)
jω
(jω)2 +5jω+6
=
jω
(jω+2)(jω+3)
Page 2 of 7
Ver 9120
E1.1 Analysis of Circuits (2016)
6. To find the low frequency asymptote, you take the lowest power of jω in each of the numerator and
denominator factors and multiply them together. Likewise, for the high frequency asymptote, you
take the highest power of jω in each of the factors. There is no need (or indeed advantage) to do
any factorization or to multiply out existing factors.
(a) HLF =
−2ω 2
1
(b) HLF =
3
1
(c) HLF =
jω×3×3
2×1×5
(d) HLF =
12
6(jω)−1
= 2(jω)2 , HHF =
= 3, HHF =
2(jω)3
4(jω)4
−2jω 3
ω4
= 2(jω)−1
= 0.5(jω)−1
= 0.9jω, HHF =
= 2jω, HHF =
jω×2(jω)6 ×5(jω)3
2×(jω)5 ×(jω)5
12
jω
=5
−1
= 12(jω)
α
7. We must have α ≥ 0 because, if α were negative, (jω) would increase without limit as ω → 0
and the transfer function would exceed G at some point. Similarly, we must have β ≤ 0 because
otherwise (jω)β would increase without limit as ω → ∞. A consequence of this is that the order of
the numerator can never exceed that of the denominator in a transfer function whose magnitude is
bounded.
8. Graphs of the transfer functions are shown in Fig. 8(a)-(c).
(a) We have a LF asymptote of 5 = 14 dB. We have corner frequencies at ω = 100 (−), 500 (+), 2000 (−)
where the sign in parentheses indicates
the polarity of gradient change. To estimate the gain at
1 + jω is equal to 1 if ω < a or else ω if ω > a. This gives
ω = 1000, we assume that
a
factor
a
a
ω/500) |H(1000j)| ' (ω5(/100
=
1
=
0
dB.
)(1)
(b) We have a LF asymptote of 2 = 6 dB. We have corner
frequencies
at ω = 100 (−), 5000 (+).
200 =
0.2 = −14 dB.
Using the same technique as in part (a), |H(1000j)| ' (ω2(1)
=
/100) 1000
(c) We have corner frequencies at ω = 100 (−), 500 (+), 2000 (−), 5000 (−). We have a LF asymptote
of 3jω = 6 dB which at the first corner
(ω = 100) is 300j = 50 dB. Using the same technique as in
3×ω(ω/500) 3000 part (a), |H(1000j)| ' (ω/100)(1)(1) = 5 = 600 = 55.6 dB. To obtain this expression from the
transfer function, any term whose corner frequency is > ω has been replaced by (1).
60
5
0
-5
|H| (dB)
0
5
|H| (dB)
|H| (dB)
10
-5
40
-10
-10
1
2
10
10
3
10
ω (rad/s)
Fig. 8(a)
4
10
-15
1
10
50
2
10
3
10
ω (rad/s)
30 1
10
4
10
2
10
Fig. 8(b)
3
10
ω (rad/s)
4
10
Fig. 8(c)
1
1
9. The corner frequency is p = ζRC
= 2π × 1000. Rearranging this gives R = 2000πζC
= 22508 Ω.
2
The upper resistor therefore has a value ζ R = 0.5R = 11254 Ω. The complete circuit is shown in
Fig. 9. At ω = 100 Hz, 1 kHz and 10 kHz the transfer function is −0.0099 + 0.0014j = 0.01∠172◦ ,
0.707j = 0.707∠90◦ and 0.9899 + 0.1414j = 1∠8◦ respectively.
0
|H| (dB)
-20
-40
-60
-80
100
Fig. 9
Solution Sheet 5
Fig. 10(a)
200
500
1k
2k
Frequency (Hz)
5k
Fig. 10(b)
Page 3 of 7
10k
Ver 9120
E1.1 Analysis of Circuits (2016)
10. For a 4th order filter, we need N = 2 and from the formula given in the question, we use ζ1 =
cos π8 = 0.924 and ζ2 = cos 3π
= 0.383. If we stick to C = 10 nF as in Q9, we obtain R1 =
8
1
1
2
2
=
17.2
kΩ
and
R
=
2
2000πζ1 C
2000πζ2 C = 41.6 kΩ with ζ1 R1 = 14.7 kΩ and ζ2 R2 = 6.1 kΩ. This gives
4
( jω
p )
the circuit shown in Fig. 10(a). The transfer function is H(jω) = jω 2
2
jω
( jω
( P ) +2ζ1 ( jω
p )+1
P ) +2ζ2 ( p )+1
; this is plotted in Fig. 10(b).
11. To shift the frequency response up by a factor of 5, we need to divide the value of each C or L
component by 5. This gives the circuit of Fig. 11(a). We could also, if we wanted, multiply all the
capacitor values by k and divide all the resistor values by k for any scale factor k without changing
the transfer function. For this particular circuit, it would be a bad idea to use a value of k > 1
because, at 3 kΩ the feedback resistor is already a little on the low side for many op-amps (which
have a limited current output capability).
To reflect the magnitude response in the line ωm = 10000, we need to convert resistors into capacitors
and vice-versa. From the notes, the formulae are: R0 = ωmk C , C 0 = ωm1kR . For the circuit of
Fig. 11(b), I have chosen k = 3.33 in order to get reasonable component values but other choices are
also possible. A full analysis of this low-pass filter circuit is the subject of question 15.
Fig. 11(a)
Fig. 11(b)
12. (a) The parallel combination of C||(R + L) has an impedance Z =
1
jωC (R+jωL)
1
jωC +R+jωL
=
R+jωL
.
1+jωRC+(jω)2 LC
We want to find the value of ω that makes this real. The easiest way to do this is to insist that the
ratio of imaginary to real part is the same for the numerator and denominator (this implies that they
ωr RC
rL
= 1−ω
from which cross multiplying (after dividing both
have the same argument). Thus ωR
2
r LC
q
2C
numerators by ωr ) gives L − ωr2 L2 C = R2 C from which ωr = L−R
= 9 950 rad/s. Note that this
L2 C
is close, but not exactly equal to, ω0 = 10 000 where the capacitor and inductor impedances have
the same magnitude. The value of Z at resonance can now be found as the ratio between the real
(or equivalently the imaginary) parts of the numerator and denominator of the previous expression.
R+jωL
jωL
Thus Z = 1+jωRC+(jω)
= 1−ωR2 LC = jωRC
= 1000.
2
LC
(b) By definition Q equals ωr times the average stored energy divided by the average power loss. If
2
the input voltage phasor is V , then the peak energy stored in the capacitor is 12 C |V | and its average
2
1
V
stored energy is half this, namely 4 C |V | . The current through the resistor is IR = R+jω
. The
rL
2
|V |2
r
C 2
R2 + L−R
L
L2 C
|
1
peak energy stored in the inductor is 12 L |IR | = 12 L R2 |V
+ω 2 L2 = 2 L
2
1
2C
2
2
2
C|V |
= 12 L R2 C+L−R
2C =
|V | which is the same as the peak capacitor energy; likewise, the average energy stored in the
2
2
2
inductor is 14 C |V | . The average power loss in the resistor is 12 R |IR | = RC
2L |V | . Calculating Q
from its definition gives Q = ωr
2
2
1
1
4 C|V | + 4 C|V |
2
RC
|V
|
2L
=
ωr L
R
= 9.95. Since the capacitor and inductor store
the same amount of energy on average, the Q can be determined more simply as Q = ωr
ωr L
R
2
1
2 L|IR |
2
1
R|I
|
R
2
=
= 9.95 .
(c) Note that the capacitor is unchanged in the two networks, so we can ignore it when matching
their impedances. When choosing components to make two networks have the same impedance,
your have a choice: you can either match their impedances or their admittances. You get the same
answer in either case, but the algebra can sometimes be much simpler in one case than the other.
In this question, it is easiest to use admittances because the components whose values are unknown
are in parallel and so their admittances add: the total admittance of RP and LP in parallel is
j
1
RP − ωr LP and RP and LP remain unentangled in this expression. The admittance of RS + LS is
Solution Sheet 5
Page 4 of 7
Ver 9120
RS −jωr LS
1
2 +ω 2 L2
RS +jωr LS = RS
r S
2
RS
+ωr2 L2S
RP =
= 1 kΩ
RS
E1.1 Analysis of Circuits (2016)
=
j
ωr LP . Equating the real and imaginary
2
R2 +ω 2 L2
LP = Sω2 LrS S = LS + ωR
= 10.1 mH.
2
r
r LS
1
RP
and
−
parts of this equation gives,
13. (a) At ω = 10 000, ZL = 100j and ZC = −100j. Therefore the currents in L and C are equal and
opposite. So the peak power supplied by V is the peak power absorbed by the resistor which equals
|V |2
R = 100 mW.
(b) The energy stored in the capacitor at time t is WC = 12 Cv(t)2 . So the power absorbed by the
dv
C
capacitor is dW
dt = Cv dt . Since you are told that the phasor V = 10, you know that the waveform
v(t) = 10 cos(ωt) and, differentiating gives dv
dt = −10ω sin(ωt). Multiplying everything out gives
=
−100ωC
cos
(ωt)
sin
(ωt)
=
−50ωC
sin
(2ωt). This has a peak value of 50ωC = 500 mW. As
Cv dv
dt
is common in resonant circuits, this is 5 times greater than the answer to part (a).
Solution Sheet 5
Page 5 of 7
Ver 9120
E1.1 Analysis of Circuits (2016)
14. (i) This is an inverting amplifier:
= − ZRF = − R1 ×
Y
X
Y
X
(ii) This is a non-inverting amplifier:
=1+
ZF
R
2R
1+2jωRC
=1+
2
= − 1+2jωRC
.
2
1+2jωRC
=
3+2jωRC
1+2jωRC .
(iii) This is the same as the previous circuit, but with an additional CR circuit at the input.
4jωRC
3+2jωRC
1
1
3
1+4jωRC × 1+2jωRC . This has corner frequencies at ωRC = 4 (−), 2 (−), 2 (+).
Y
X
=
(iv) The circuit has negative feedback so we can assume V+ = V− = 0. KCL @ V− gives: 0−X
R +
0−Y
0−Z
Z−Y
Z
R + R = 0 from which −Z = X + Y . Now KCL @ Z gives: (Z − 0)jωC + R + R = 0 from
which Y − Z (2 + jωRC) = 0. Substituting −Z = X + Y gives Y + (X + Y ) (2 + jωRC) = 0 from
Y
= − 2+jωRC
which X
3+jωRC .
10
5
8
|H| (dB)
|H| (dB)
0
-5
-10
6
4
2
0
-15
0.1/RC
ω (rad/s)
1/RC
0.1/RC
Fig. 14(i)
10/RC
Fig. 14(ii)
10
0
5
-1
|H| (dB)
|H| (dB)
1/RC
ω (rad/s)
0
-2
-5
-3
-10
0.1/RC
1/RC
ω (rad/s)
10/RC
1/RC
Fig. 14(iii)
10/RC
ω (rad/s)
Fig. 14(iv)
15. In this circuit, the output, Y , is fed back to both V+ and V− so it is not immediately obvious that
the overall feedback is negative. However, we see that V− = Y whereas |V+ | will be attenuated by
the network and will be < Y , so all is well. We can therefore assume that Z = V+ = V− = Y .
(a)
Z
W
is just a potential divider, so
Z
W
=
Y
W
=
1
jωC1
1
R2 + jωC
1
1+jωR2 C1 .
=
1
Y = Z as noted above. From
this we get W = Y (1 + jωR2 C1 ).
(b) KCL @ W gives:
W −X
R1
+
W −Z
R2
+ (W − Y ) jωC2 = 0 from which (substituting Z = Y ),
W (R1 + R2 + jωR1 R2 C2 ) − Y (R1 + jωR1 R2 C2 ) − XR2 = 0.
Substituting the expression for W above gives
Y (1 + jωR2 C1 ) (R1 + R2 + jωR1 R2 C2 ) − Y (R1 + jωR1 R2 C2 ) = XR2
2
from which Y R2 + jωR2 (R1 + R2 ) C1 + (jω) R1 R22 C1 C2 = XR2 .
Y
X
Hence
(jω) =
1
.
R1 R2 C1 C2 (jω)2 +(R1 +R2 )C1 jω+1
2
2 2
p
(R
1 +R2 ) C1
expression for ζ gives ζ 2 =
4
(R +R )2 C 2
C1
1 R2
(c) Squaring the
= 4R11 R22C1 C21 which gives (R4R+R
2 = ζ2C .
2
1
2)
Using quite a common algebraic trick, we can write the numerator as the difference of two squares:
2
2
2
(R1 +R2 )2 −(R1 −R2 )2
R1 −R2
2R1
4R1 R2
2
=
= 1 − R1 +R2 = 1 − R1 +R2 − 1 = 1 −
.
R2 − 1
(R +R )2
(R +R )2
1
2
Rearranging
1
C1
ζ 2 C2
1+ R
2
1
=1−
2
R
1+ R2
1
1+
q
R2
1
+
1−
R1
C1
ζ 2 C2
2
−1
gives
2
R
1+ R2
=1+
q
1−
C1
ζ 2 C2
from which
1
= 2.
R2
The usefulness of this relationship is that it allows you to determine the resistor ratio, R
, if you
1
C2
C1
know the capacitor ratio C1 . For the square root to be a real number, we must have 1 − ζ 2 C2 ≥ 0
which implies C1 ≤ ζ 2 C2 .
1
2
(d) We must have C
C1 ≥ ζ 2 = 4. Given our restricted choice of capacitor value, we must therefore
1
choose C2 = 47 nF and C1 = 10 nF. So, substituting ζ C
= 0.851 into the expression from the
2C
2
Solution Sheet 5
Page 6 of 7
Ver 9120
E1.1 Analysis of Circuits (2016)
previous part, we find
1+
R2
R1
p2 , we can write 0.443R12 = R1 R2 =
R2 = 0.443R1 = 4.9 kΩ.
1
p2 C1 C2
= 0.443. From the expression for
q
53.9×106
= 53.9 × 106 . Hence R1 =
= 11 kΩ and
0.443
0
0
-0.5
arg(H) (rad)
-10
|H| (dB)
R2
R1
× 1.386 = 2 from which
-20
-30
-1
-1.5
-2
-2.5
-3
-40
0.1/RC
1/RC
ω (rad/s)
0.1/RC
1/RC
ω (rad/s)
Fig. 15(mag)
Fig. 15(phase)
16. (a) This circuit is a potential divider, so (setting R = 20) we can write down the transfer function:
q
2ζ ( jω
4jωRC
a )
Y
4R
1
=
=
where
a
=
=
1
jω 2
X
LC = 5000 and ζ = 2.5aRC =
5R+jωL+ jωC
1+5jωRC+(jω)2 LC
1+2ζ ( jω
a )+( a )
0.1.
Y it is easiest to find instead the maximum of Y 2 = Y ×Y ∗∗ where
(b) To find the maximum of X
X
X×X
the * denotes the complex conjugate. Note that (i) a number multiplied by its complex conjugate
is just the sum of the squares of its real and imaginary parts and that (ii) the magnitude of a
complex fraction is the magnitude of the numerator divided by the magnitude of the denominator;
very rarely is it necessary to multiply the top and bottom of a fraction by the complex conjugate of
the denominator.
2
Y 2 4jωRC
=
The difficult way to find the maximum is to differentiate the expression X
=
2
1+5jωRC+(jω) LC
(4ωRC)2
and
(1−ω 2 LC)2 +(5ωRC)2
2
Y 2
4R
= 1 X
5R+jωL+ jωC
q
set the derivative to zero. Much easier is to take the first expression above:
16R2
1
=
ωL − ωC
= 0
2 . This is clearly maximized by making
1
25R2 +(ωL− ωC
)
Y
At this frequency X
= 0.8 = −1.9 dB.
Y 2
1 2
(c) The 3dB bandwidth is when X has fallen by a factor of 2. This will happen when ωL − ωC
=
1
25R2 or ωL − ωC
= ±5R. So we need to solve the quadratic equation LCω 2 ± 5RCω − 1 = 0. The
√
√
25R2 C 2 +4LC
25R2 C 2 +4LC
of which the positive solutions are ω = ±5RC+ 2LC
.
solution is ω = ±5RC± 2LC
This gives ω3dB = {4525, 5525}. The bandwidth is the difference between these which is 10RC
=
LC
1000 rad/s. Notice that ω0 is the geometric mean of the two 3dB frequencies but is not the arithmetic
1
mean which is 5025 rad/s. The Q (quality factor) of the resonance is Q = 2ζ
= 5. This also equals
the ratio of ω0 to the bandwidth and the height of the peak above the intersection of the asymptotes.
which means ω0 =
1
LC .
The circles in Fig. 16 indicate ω0 and the two 3dB frequencies.
1
arg(H) (rad)
|H| (dB)
-10
-20
-1
-30
3
10
4
ω (rad/s)
10
Fig. 16(mag)
Solution Sheet 5
0
3
10
4
ω (rad/s)
10
Fig. 16(phase)
Page 7 of 7
Ver 8954
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 6 (Lectures 14, 15 & 16)
Key: [A]= easy ... [E]=hard
Note: A tilde-superscript on a phasor denotes division by
equals the RMS value of a phasor V .
√
2, i.e. Ve =
√1 V
2
. This means that Ve 1. [A] Say which of the following waveforms include negative exponentials and which include positive
exponentials: (a) 2 − 4e−3t , (b) 2 + 4e3t , (c) 2 + 4e−3t , (d) −2 − 4e3t , (e) 2 + 4e−t/−3 .
2. [B] Suppose v(t) = 5 + 2e−100t .
(a) Determine the time constant, τ , of the negative exponential.
(b) Determine the time at which v(t) = 5.5 V.
(c) Give an expression for the time taken for v(t) to fall from A to B where 5 < B < A < 7.
3. [B] If V = −200j in Fig. 3, find the phasor value of I and the complex power absorbed by each of
the components including the voltage source.
(
0 t<0
in Fig. 4 (below),
4. [B] If v(t) =
5 t≥0
(a) find an expression for x(t) for t ≥ 0.
(b) Sketch a graph of x(t) for −RC ≤ t ≤ 3RC.
(c) Determine the time at which x(t) = 4.5.
5. [C] For each of the circuits shown in Fig. 5(i)-(vi)
determine (a) the time constant (b) the DC gain
Y Y and
(c)
the
high
frequency
gain
.
In
each case, determine these in two ways: directly
X ω=0
X ω=∞
from the circuit and via the transfer function.
Fig. 3
Fig. 5(i)
Fig. 5(ii)
Fig. 5(iii)
Fig. 4
Fig. 5(iv)
Fig. 5(v)
Fig. 5(vi)
6. [C] For each of the periodic waveforms shown in Fig. 6(i)-(iii) determine (a) the mean value and (b)
the rms value.
6
4
4
3
2
2
0
1
-2
0
1
2
3
Fig. 6(i)
Problem Sheet 6
4
5
6
0
0
6
4
2
1
2
3
4
Fig. 6(ii)
5
6
0
0
1
2
3
4
5
6
Fig. 6(iii)
Page 1 of 3
Ver 8954
E1.1 Analysis of Circuits (2016)
(
0
7. [C] If v(t) =
5
t<0
in Fig. 7, determine an expression for x(t) for t ≥ 0 and sketch its graph.
t≥0
Fig. 7
(
8. [C] If v(t) =
Fig. 8
Fig. 9
Fig. 10
2 t<0
in Fig. 8, determine an expression for x(t) for t ≥ 0 and sketch its graph.
6 t≥0
(
4 t<0
in Fig. 9, find an expression for x(t) for t ≥ 0 and sketch a graph of x(t) for
9. [C] If v(t) =
1 t≥0
the time interval −RC ≤ t ≤ 3RC.
10. [C] Fig. 10 shows a simplified circuit diagram for an oscilloscope probe which includes an adjustable
capacitor of value kC.
(a) Determine the transfer function,
Y
X (jω)
and determine its value at ω = 0 and ω = ∞.
(b) Determine the time constant of the circuit.
(
0
(c) Determine an expression for y(t) if x(t) =
10
t<0
.
t≥0
(d) The variable capacitance, kC, is adjusted to the value that makes the amplitude of the transient
equal to zero. Determine the value of k that achieves this.
(e) Simplify the expression for
Y
X (jω)
when k has the value calculated in the previous part.
11. [C] In the diagram of Fig. 11 power is being transmitted from a source to a load via two transformers
having turns ratios of 1 : n and n : 1 respectively.
(a) If VeL = 240 V and the average power dissipated in RL is 10 kW, calculate the value of RL .
(b) If RS = 0.5 Ω, calculate the power dissipated in RS when (i) n = 1 and (ii) n = 5.
Fig. 11
Fig. 12
12. [C] The circuit in Fig. 12 represents a microphone connected to an amplifier via a transformer and
a long cable.
(a) Determine the Thévenin output impedance of the microphone+transformer combination when
n = 4.
(b) The cable is subject to 50 Hz interference capacitively coupled from the mains, VeN = 230 V, via
a capacitor of value 100 pF. If the RMS microphone signal amplitude is VeS = 1 V, calculate the
ratio of the signal and the noise at the amplifier in dB if (i) n = 1 and (ii) n = 4.
Problem Sheet 6
Page 2 of 3
Ver 8954
E1.1 Analysis of Circuits (2016)
13. [C] In the circuit of Fig. 13, the transformer may be assumed to be ideal.
(a) Calculate the average power dissipated in each of R1 and R2 if Ves = 1, n1 = 2, n2 = 3, R1 = 10
and R2 = 20.
(b) Calculate, in terms of n1 , n2 , R1 and R2 , the effective resistance seen by the voltage source.
Fig. 13
Fig. 14
14. [C] In the circuit of Fig. 14, VeS = 230 at 50 Hz, L = 8 mH and R = 1.6 Ω.
(a) If C = 0 (i.e. the capacitor is omitted), calculate the apparent power, average power and
reactive power absorbed by the load (shaded region) and also its power factor.
(b) Determine the value of C needed to increase the power factor to 0.9. Using this value, recalculate
the quantities from part (a).
15. [D] Calculate the waveform y(t) in Fig. 15(i) when,
(
0
t≤0
(a) v(t) =
as shown in Fig. 15(ii).
5 sin 2000πt t > 0


t≤0
0
(b) v(t) = 5 sin 2000πt 0 < t ≤ 1 ms as shown in Fig. 15(iii)


0
t > 1 ms
5
5
0
0
-5
-1
0
1
2
-5
-1
0
t (ms)
Fig. 15(i)


open
16. [D] If the switch in Fig. 16 is closed


open
periods and sketch graphs of i(t) and
1
2
t (ms)
Fig. 15(ii)
Fig. 15(iii)
t<0
0 ≤ t < 2 ms , determine expressions for i(t) for each of these
t ≥ 2 ms
v(t) for −1 ms ≤ t ≤ 4 ms.
3
2
1
0
-1
Fig. 16
Fig. 17(i)
0
1
t (ms)
2
3
Fig. 17(ii)


0 t < 0
17. [E] In Fig. 17(i), v(t) = 3 0 ≤ t < 1 ms as shown in Fig. 17(ii). If the diode has a forward voltage


2 t ≥ 1 ms
drop of 0.7 V, find expressions for x(t) for t ≥ 0.
Problem Sheet 6
Page 3 of 3
Ver 8952
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 6 - Solutions
1. (a) Negative, (b) Positive, (c) Negative, (d) Positive, (e) Positive.
1
2. (a) The time constant is 100
= 10 ms. (b) We need to solve 5.5 = 5 + 2e−100t ⇒ e−100t = 0.25 ⇒
−100t = ln 0.25 = −1.386
⇒ t = 13.86 ms. Alternatively, we can use the standard formula, derived
7−5
in lectures, t = τ ln 5.5−5
= 10 × ln 4 = 13.86 ms. (iii) The general formula, derived in lectures, is
A−5
TA→B = τ ln B−5
.
2
2
= −24 − 32j = 40∠ − 127◦ . So Ie = 402 = 800. The
2
2
complex power absorbed by each of the passive components is Ie Z; this gives Ie R = 3.2 kW,
2
2
e
I ZL = 4000j = 4 kVAR and Ie ZC = −1600j = −1.6 kVAR. The current through the source
(following
∗ the passive sign convention) is −I = 24 + 32j and the complex power absorbed by it is
Ve −Ie = −141j (17 − 22.6j) = (−3.2 − 2.4j) kVA. As expected, the total complex power sums to
zero.
(
0 t<0
. Because x
4. (a) The DC gain of the circuit is 1, so the steady state output is xSS (t) =
5 t≥0
is the voltage across a capacitor, it must be continuous, so x(0+) = x(0−) = 0. So the complete
−t
−t
expression is x(t) = xSS (t) + (x(0+) − xSS (0+)) e τ = 5 − 5e τ where τ = RC. (b) x(t) is plotted
in Fig. 4.
0−5
(c) Using the standard formula, t = τ ln 4.5−5
= τ ln (10) = 2.3RC.
3. The current is I =
−200j
4+5j−2j
=
−200j
4+3j
4
2
0
-1
0
1
t/RC
2
3
Fig. 4
5. (i) From the circuit, the time constant is RC = 2 ms. The DC gain may be obtained by treating C as
an open circuit and is 0; the HF gain may be obtained by treating C as a short-circuit and is therefore
jωRC
Y
1. The transfer function (using potential divider formula) is X
(jω) = 1+jωRC
which happily gives
Y
Y
1
the same values: X (0) = 0, X (∞) = 1, τ = denominator corner frequency = RC. The importance of
Y
X
(∞) is that it gives the gain for a step input discontinuity, i.e.
Y
X
(∞) =
output discontinuity
input discontinuity .
(ii) From the circuit: τ = RC = 2 ms, DC gain (C open-circuit) = 1, HF gain (C short-circuit) = 0.
Y
1
The transfer function (using potential divider formula) is X
(jω) = 1+jωRC
which gives the same
values.
L
(iii) From the circuit: τ = R
= 0.1 ms, DC gain (L short-circuit) = 0, HF gain (L open-circuit) = 1.
jωL
Y
The transfer function (using potential divider formula) is X
(jω) = R+jωL
which gives the same
values. Note that if the denominator is (p + jωq), the time constant is pq and the corner frequency
is pq .
(iv) To obtain the time constant, we need to determine the Thévenin resistance seen by the inductor.
To do this, we set the input voltage, X, to zero (thereby shorting node X to ground) and find
the resistance between the inductor terminals (with the inductor removed). The two resistors are
L
in series, so we get RT h = 1 + 4 = 5 k. From this, τ = R
= 200 ns, DC gain (L short-circuit)
= 0.2 (potential divider), HF gain (L open-circuit) = 1. The transfer function (using potential
R+jωL
Y
divider formula) is X
(jω) = 5R+jωL
where R = 1 k. This gives the same values and is, perhaps, a
marginally easier way to determine them.
Solution Sheet 6
Page 1 of 6
Ver 8952
E1.1 Analysis of Circuits (2016)
(v) To determine the Thévenin resistance seen by the capacitor, we set X = 0 and measure the
resistance at the capacitor terminals. Since X is connected to ground, the two 1 k resistors are in
series and so we have 8 k in parallel with 2 k which gives 1.6 k. From this, τ = RC = 0.16 ms, DC gain
(C open-circuit) = 0.1 (potential divider), HF gain (C short-circuit) = 0.5. The transfer function
1+8jωRC
Y
(jω) = 2R+ R 8R
= 10+16jωRC
where R = 1 k and we used
(using potential divider formula) is X
1+8jωRC
the formula for Z8R||C =
8R
1+jω8RC .
(vi) To determine the Thévenin resistance at the capacitor terminals is not trivial because of the
dependent voltage source that is the opamp. If we set X = 0 and replace the capacitor with a voltage
source V as in Fig. 5(i), we can use nodal analysis to determine I and then calculate RT h = VI . Since
W −V
W −V
=0⇒W =
the op-amp is a unit-gain buffer, Y = V . KCL at node W gives: W
10 + 10 + 10
1
V
V
⇒ I = V −W
== 310 = 30
⇒ RT h = 30 k. So, finally, we get τ = RT h C = 3 ms. For the DC
10
gain, we make C an open-circuit as in Fig. 5(ii). Negative feedback means V+ = Y and, since there
is no current through the resistor connected to V+ , we must also have W = Y . KCL at node W
then gives W = Y = X, so the DC gain is 1. For the HF gain, the capacitor acts a a short circuit
so V+ = 0 which in turn means that Y = 0 so the gain is 0.
2
3V
Rather easier is the transfer function approach. We know V+ = Y and V+ is determined from W
1
−X
−Y
−Y
Y
= VW+ = 1+jωRC
. KCL at W gives W R
+ WR
+ WR
=
by an RC potential divider giving: W
Y
1
0 ⇒ 3W − 2Y = X. We now substitute for W using the previous equation W = 1+jωRC to get
Y
1
Y
3Y (1 + jωRC) − 2Y = X⇒ X
= 1+3jωRC
. From this we can easily get: τ = 3RC, X
(0) = 1 and
Y
X (∞) = 0.
4
2
0
-1
Fig. 5(i)
Fig. 5(ii)
0
1
t*R/L
2
3
Fig. 7
6. (i) v(t) equals 6 for 13 of the time and −2 for 32 of the time. So its average value is v = 13 ×6+ 23 ×(−2) =
√
1
2
2
2
2
2
14.67 = 3.83. This is
3 . Similarly, the average value of v is v = 3 × 36+ 3 × 4 = 14 3 . So Vrms =
higher than the average value v.
(ii) During the first period (0 ≤ t ≤ 2), the formula for v can be derived as v = 2t. To find
the average value, we integrate over one period, and divide by the length of the period. So v =
2 2
R
1 2
1
2 t=0 2tdt = 2 t 0 = 2. This is also pretty obvious from looking at the waveform. In the same
√
2
R2
2
way, v 2 = 12 t=0 (2t) dt = 12 34 t3 0 = 5 31 giving Vrms = 5.33 = 2.31.
(iii) This is the same as the previous waveform but shifted up by +2. You can perform integrations
similar to the previous part or, easier, just modify the previous answers. v (which previously equalled
2) will be increased by 2 to become v = 4. Adding a constant onto a random variable does not affect
2
2
its variance, so v 2 − (v) will be unchanged at 5 13 − 22 = 1 13 . It follows that v 2 = (v) + 1 31 = 17 13 .
√
Taking the square root gives Vrms = 17.33 = 4.16.
7. Method 1 (inductor current continuity): For t < 0, x = 0 and the current through the inductor
is i = v−x
R = 0. It follows that at time t = 0+, the current through the resistor (which equals
the current through the inductor) will still be zero and x(0+) = v(0+) = 5. From this value is
will decay to a steady state value xSS = 0 since the inductor is a short circuit for DC. Thus,
−t
−t
L
x(t) = xSS (t) + (x(0+) − xSS (0+)) e τ = 0 + 5e τ where τ = R
. This is plotted in Fig. 7.
Method 2 (transfer function): The transfer function of the circuit is (from potential divider equation)
jωL
X
V = R+jωL . From this we get the DC gain, GDC = 0, the HF gain, GHF = 1, and the time constant
L
is R
. The DC gain allows us to calculate the steady state xSS (t) = GDC v(t) ≡ 0. The output
discontinuity at t = 0 is given by ∆x = GHF ∆v = 1 × 5 = 5. So x(0+) = xSS (0−) + ∆x = 0 + 5 = 5.
−t
−t
Finally we put everything together to get: x(t) = xSS (t)+(x(0+) − xSS (0+)) e τ = 0+(5 − 0) e τ .
Solution Sheet 6
Page 2 of 6
Ver 8952
E1.1 Analysis of Circuits (2016)
8. Method 1 (capacitor voltage continuity): The time constant of the circuit is obtained by setting
v = 0 and finding the Thévenin resistance across the capacitor terminals. Since v is connected to
ground, the two resistors are (in parallel and RT h = 12 R giving τ = 12 RC. The DC gain of the circuit
1 t<0
. For t < 0, the capacitor voltage is v − x = 2 − 1 = 1. This
3 t≥0
must remain continuous and so v(0+) − x(0+) = 1 ⇒ x(0+) = v(0+) − 1 = 5. Putting everything
−t
−t
−t
together, we get x(t) = xSS (t) + (x(0+) − xSS (0+)) e τ = 3 + (5 − 3) e τ = 3 + 2e τ . This is
plotted in Fig. 8.
is 0.5, so xSS (t) = 0.5v(t) =
Method 2 (transfer function): The transfer function of the circuit is (from potential divider equation)
R
X
= 1+jωRC
R
V = R+
2+jωRC . From this we get the DC gain, GDC = 0.5, the HF gain, GHF = 1, and
1+jωRC
the time constant is 0.5RC. The DC gain allows us to calculate the steady state as above. The output
discontinuity at t = 0 is given by ∆x = GHF ∆v = 1 × 4 = 4. So x(0+) = xSS (0−) + ∆x = 1 + 4 = 5.
−t
−t
Finally we put everything together to get: x(t) = xSS (t)+(x(0+) − xSS (0+)) e τ = 3+(5 − 3) e τ =
−t
3 + 2e τ .
8
8
x
6
6
x
4
4
xSS
2
0
-1
0
1
t/RC
2
xSS
2
xTr
0
-1
3
Fig. 8
xTr
0
1
t/RC
2
3
Fig. 9
9. For opamp circuits, it is easiest to use the transfer function to determine the relevant circuit paraR
1+jωRC
meters. This is a non-inverting amplifier with a gain of X
= 2+jωRC
V = 1 +
R
1+jωRC . Thus we
have a DC gain, GDC = 2, a high frequency gain GHF = 1 and a time constant τ = RC.
At t = 0, the input discontinuity is ∆V = v(0+) − v(0−) = −3 and so the ouput discontinuity is ∆X = x(0+) −
( x(0−) = GHF ∆V = 1 × −3 = −3. The steady state output is given by
8 t<0
xSS (t) = GDC v(t) =
. So this gives x(0+) = −3 + x(0−) = −3 + 8 = 5.
2 t≥0
Putting everything together, we get x(t) = xSS (t) + (x(0+) − xSS (0+)) e
−t
2 + 3e RC .
10. (a) The impedance of R||C is
gain is
Y
X
= H(jω) =
R
1+jωRC .
R
1+jωRC
9R
R
1+j9kωRC + 1+jωRC
−t
τ
= 2 + (5 − 2) e
−t
τ
=
We can regard the circuit as a potential divider and so the
which simplifies to H(jω) =
H(0) = 0.1 and the high frequency gain is H(∞) =
1+j9kωRC
10+j9(k+1)ωRC .
The DC gain is
k
k+1 .
(b) The time constant is the reciprocal of the denominator corner frequency and is τ = 0.9(k + 1)RC.
(c) The steady state output for an input of x = 10 is ySS (t) = H(0) × 10 = 1. The input step is
10k
∆x = 10 and so the output step is ∆y = H(∞)∆x = k+1
. Since y(t) = 0 for t < 0, it follows that
t
10k
10k
− τt
y(0+) = k+1 . So the output for t > 0 is y(t) = 1 + k+1 − 1 e− τ = 1 + 9k−1
.
k+1 e
(d) The value of k that makes the transient amplitude zero is k = 19 . Note that the transient
amplitude will be positive or negative according to whether k is greater or less than this value.
(e) If k =
1
9
then H(jω) = 0.1 and is independent of ω.
e2
V
RL ,
e
RS is InL
11. (a) Average power is
current through
so RL =
e2
V
10 k
= 5.76 Ω. (b) Current through RL is IeL =
so the power dissipation is
2
IeL
RS
n2
e
V
R
= 41.7 A. The
. This gives (i) 868 W and (ii) 34.8 W.
12. (a) Impedances are transformed by the square of the turns ratio because the voltage decreases by n
and the current increases by n so that the ratio of voltage over current decreases by n2 . So when
n = 4, the impedance at the output of the secondary is 2400
16 = 150 Ω.
Solution Sheet 6
Page 3 of 6
Ver 8952
E1.1 Analysis of Circuits (2016)
1
(b) At 50 Hz, the capacitor impedance is ZC = j2π50×100
p = −j31.8 MΩ . With the transformed
source impedance from part (a), we get the equivalent circuit shown in Fig. 12. Using superposition,
−2
eS +Rn−2 V
eN
ZC
n−1 ZC V
VeA = VeN × RnRn
+ VeS n−1 × Rn−2
. The ratio of the signal and noise
−2 +Z
+ZC =
Rn−2 +ZC
C
voltages is equal to the ratio of the two terms in the numerator, so the ratio of the signal and noise
eS |
n−1 |ZC ||V
31.8 M×1
voltage magnitudes is Rn−2 Ve
× n = 57.66 × n. Converting this to decibels gives
= 2400×230
| N|
20 log10 57.66 + 20 log10 n = 35.2 + 20 log10 n. This gives (i) 35.2 dB for n = 1 and (ii) 47.3 dB for
n = 4. So, slightly surprisingly, using a transfomer to reduce the voltage coming from the microphone
actually makes the signal-to-noise ratio better.
Fig. 12
Fig. 13
13. (a) Ve1 = n1 VeS = 2 so the average power dissipated in R1 is
so the average power dissipated in R2 is
e2
V
2
R2
e2
V
1
R1
= 400 mW. Similarly, Ve2 = n2 VeS = 3,
= 450 mW.
(b) From the ideal transformer equations, 1 × IS + n1 × (−I1 ) + n2 × (−I2 ) = 0 (the minus signs arise
because in Fig. 13 I1 and I2 are defined as coming out of the transformer). Rearranging this and
n21 VS
n22 VS
1
using Ohm’s law gives IS = n1 I1 +n2 I2 = nR1 V1 1 + nR2 V2 2 = R
+ R
. So Ref f = VISS = 1 +
.
1
1
2
−2
n1 R1
−2
n2 R2
−2
This is the same as the parallel combination of n−2
1 R1 and n2 R2 , i.e. the parallel combination of
the individual winding resistances transferred from the secondaries to the primary.
VS
=
14. (a) The inductor impedance is jωL = j × 100π × 0.008 = 2.51 Ω. So the current is I = R+jωL
∗
41.5−65.1j. So S = V I = P +jQ = 9.54+14.98j kVA. So the apparent power is |S| = 17.8 kVAand
P
the average and reactive powers are P and Q given earlier. The power factor is cos φ = |S|
= 0.54.
(b) Adding the capacitor will not consume any average power and so will not affect P at all. We
need to reduce Q to P tan (arccos 0.9) = 4.62 kVAR since tan φ = Q
P and we want cos φ = 0.9. It
2
S|
2
follows that QC = 4.62 − 14.98 = −10.36 kVAR = −|V
|ZC | = −230 ωC. This gives C = 623 µF. Now
S = P + jQ = 9.54 + 4.62j kVA. So the apparent power is |S| = 10.6 kVA and the average and
P
reactive powers are P and Q given earlier. The power factor is cos φ = |S|
= 0.9.
15. Notice first that since there are no input discontinuities, there will be no output discontinuities
jωRC
either. The circuit transfer function is X
V = 1+jωRC . The gain at ω = 2000π is G = 0.503 +
0.5j = 0.709∠44.8◦ . The phasor corresponding to v(t) = 5 sin ωt is V = −5j and so the steady
state output will be X = GV = 2.5 − 2.51j = 3.54∠ − 45.2◦ which corresponds to a waveform
xSS (t) = 2.5 cos ωt + 2.51 sin ωt. The time constant is RC = 0.16 ms.
(a) At time t = 0+ we have xSS (0+) = 2.5 V. Since there is no output discontinuity, x(0+) =
−t
x(0−) = 0. Putting this together gives x(t) = xSS (t) + (x(0+) − xSS (0+)) e τ = 2.5 cos ωt +
−t
2.51 sin ωt + (0 − 2.5) e RC . This is plotted in Fig. 15(i).
(b) Substituting t = 1 into the previous expression gives x(1−) = x(1+) = 2.495 (very close to the
steady state value since it has had 6 14 time constants to converge). For t > 1, the steady state is
xSS (t) ≡ 0. Therefore we get x(t) = 2.495e
Solution Sheet 6
−(t−1)
τ
. This is plotted in Fig. 15(ii).
Page 4 of 6
Ver 8952
E1.1 Analysis of Circuits (2016)
5
5
xSS
xSS
x
0
0
xTr
-5
-1
xTr
0
1
-5
-1
2
x
0
1
t (ms)
2
t (ms)
Fig. 15(i)
Fig. 15(ii)
16. For the steady state when t < 0, we can treat the inductor as a short circuit and so i = 10
R = 100 mA.
V
di
When the switch is closed, there is a constant 10 V across the inductor and so dt = L = 100 V/s.
Therefore the current through the inductor will increase linearly at this rate for 2 ms (from an initial
value of 100 mA) and will reach a value of 300 mA. Since there is no resistor in series with the
inductor, the current increases linearly rather than exponentially; you can, if you wish, regard this
as a limiting case of a negative exponential that has an infinite time constant.
When the switch is opened at t = 2 ms, the current will decay from its peak value of 300 mA back
L
down to its steady state value of 100 mA with a time constant of R
= 1 ms. Thus for t > 2 ms, we
have i(t) = 100 + (300 − 100) e
−(t−2)
1
(in units of mA and ms). All this is plotted in Fig. 16(i).
When the switch is open, v(t) = Ri(t). However, when the switch is closed, v(t) ≡ 0. We therefore
get the voltage waveform plotted in Fig. 16(v).
300
30
v
i
200
100
0
-1
0
iTr
20
iSS
10
1
2
3
vSS
0
-1
4
0
t (ms)
1
2
3
4
t (ms)
Fig. 16(i)
Fig. 16(v)
17. In this question, we have two different circuits according to whether the diode is off or on. These are
shown in Fig. 17(off),(on). When the diode is off, we have a DC steady state xSS (t) = 0 and a time
constant τOf f = RC = 1.6 ms. On the other hand, when the diode is on, we can get the DC steady
state by doing KCL for the shaded supernode: x+0.7−v
+ x8 = 0 ⇒ xSS = 45 (v − 0.7). We obtain
2
the time constant by setting all voltage sources to zero and finding the Thévenin resistance and the
capacitor terminals: this is RT h = 2 k||8 k = 1.6 k; this gives a time constant τOn = 0.32 ms.
For t < 0, v = x = 0 and so, since v − x < 0.7, the diode will be off. When v changes to 3, the diode
will turn on and will charge the capacitor up to a steady state voltage of 45 (3 − 0.7) = 1.84. When
v now changes to 2 V, the diode will turn off and x will fall towards the “off” steady state of 0 V.
However, it will never reach this value, because when x reaches v − 0.7 = 1.3 V the diode will turn
on again resulting in a new steady state of 45 (2 − 0.7) = 1.04 V. So this means we actually have four
distinct time segments: t < 0, 0 ≤ t < 1, 1 ≤ t < Tx , t ≥ Tx wher Tx is the, as yet unknown, time
at which the diode turns on for the last time.
Segment 1 (Diode Off, t < 0, x = v = 0).
Segment 2 (Diode On, 0 ≤ t < 1, v = 3, xSS =
xSS (t) + (x(0+) − xSS (0+)) e
x(1) = 1.76.
−t
τ
4
5
= 1.84 + (0 − 1.84) e
(3 − 0.7) = 1.84, τOn = 0.32 ms): x(t) =
−t
τ
−t
= 1.84 − 1.84e τOn . At t = 1 this gives
Segment 3 (Diode Off, 1 ≤ t < Tx , v = 2, xSS = 0, τOf f = 1.6 ms): Capacitor voltage continuity
−t
−t
means that x(1+) = 1.76. So x(t) = xSS (t) + (x(1+) − xSS (1+)) e τ = 0 + (1.76 − 0) e τ =
1.76e
−(t−1)
τOf f
. We need to know when the voltage x reaches 1.3 V (t = Tx ) because that is when the
diode will turn on again. Solving 1.76e
−(Tx −1)
τOf f
= 1.3 ⇒ Tx = 1.48 ms.
Segment 4 (Diode On, t ≥ 1.48 ms, v = 2, xSS =
xSS (t) + (x(Tx +) − xSS (Tx +)) e
−t
τ
4
5
(2 − 0.7) = 1.04, τOn = 0.32 ms): x(t) =
= 1.04 + (1.3 − 1.04) e
−t
τ
= 1.04 + 0.26e
−(t−Tx )
τOn
.
All four segments are plotted in Fig. 17(iii).
Solution Sheet 6
Page 5 of 6
Ver 8952
E1.1 Analysis of Circuits (2016)
2
xSS
x
xSS
0
xTr
xTr
-2
-1
Fig. 17(off)
Solution Sheet 6
Fig. 17(on)
0
1
t (ms)
2
Fig. 17(iii)
Page 6 of 6
3
Ver 8284
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 7 (Lectures 17 & 18)
Key: [A]= easy ... [E]=hard
Note: In this problem sheet u and Z0 are the propagation velocity and characteristic impedance
of a transmission line and the forward and backward waves at the point x are fx (t) = f (t − ux ) and
gx (t) = g(t + ux ) with, in the case of sinusoidal waves, the corresponding phasors being Fx and Gx .
[A] Find the propagation velocity, u, and characteristic impedance, Z0 , of a transmission line
whose capacitance and inductance are 50 pf/m and 500 nH/m respectively. Express the propagation velocity also as a fraction of the speed of light. [B] The line in Fig. has Z0 = 100 Ω. For
each of the cases below, calculate the reflection coefficients at both ends of the line and describe
the waves that would arise from a short positive pulse at VS .RS = 10 and RL = 100.RS = 10
and RL = 1000.RS = 100 and RL = 1000.
(b)
(a)
(c)
Fig.
Fig.
[C] The line in Fig. is driven by a 10 V DC voltage source. Determine the voltage and current in
the line and hence the forward and backward waves f (t − ux ) and g(t + ux ). Determine also the power
carried by the two waves and verify that their differnce equals the total power delivered to the load.
Fig.
Fig.
[C] A transmission line has a propagation velocity of 15 cm/ns and a characteristic impedance of
100 Ω. The forward and backward waves are shown in Fig. and have amplitudes of 9 V and 3 V
respectively. Draw dimensioned sketches of the voltage and current waveforms at (a) x = 0 and
(b) x = 300 cm. In each case, give the value of the peak voltage and peak current.
[C] The transmission line shown in Fig. has a propagation velocity of 15 cm/ns and a characteristic
impedance of 50 Ω. The length of the line is L = 300 cm.
(a) Determine the reflection coefficients at both ends of the line when the switch is held closed.
(b) Calculate the steady state DC forward and backward waves when the switch has been closed
for a long time.
(c) If the switch is closed at time t = 0, determine the forward and backward waves at x = 0. Hence
determine the voltage waveforms at x = 0 and x = L.
[C] A length of transmission line with Z0 = 100 and u = 20 cm/ns is terminated in a short circuit at
x = L. Find the shortest lengths of line, L, for which the impedance at 20 MHz at x = 0 will equal
(a) 50 pF and (b) 1 µH.
Problem Sheet 7
Page 1 of 2
Ver 8284
E1.1 Analysis of Circuits (2016)
[C] In Fig. , L = 5 m, u = 20 cm/ns, Z0 = 100 , RL = 50 and the frequency of operation is 50 MHz.
(a) If the forward wave phasor at x = 0 is F0 = 6j, determine the forward wave phasors, Fx , at
x = 1, 2, 3, 4 and 5 metres.
(b) Calculate the reflection coefficient at x = L.
(c) Determine the backward wave phasors, Gx , at x = 0, 1, 2, 3, 4 and 5.
(d) Determine the line voltage phasors, Vx , at x = 0, 1, 2, 3, 4 and 5.
(e) Determine the Voltage Standing Wave Ratio: V SW R =
(f) Determine the line impedance,
V0
I0 ,
max(|Vx |)
min(|Vx |) .
at x = 0.
Fig.
Fig.
[C] Repeat question for RL = 100.
[C] In Fig. , L = 1 m, u = 15 cm/ns, Z0 = 100 , RS = 10 , RL = 150 and the frequency of operation
is 20 MHz.
(a) Calculate the reflection coefficient, ρL at x = L. Hence calculate the phasor ratio
(b) Calculate the line impedance
V0
I0
G0
F0
at x = 0.
at x = 0.
(c) By treating the circuit at the source end as a potential divider, calculate V0 if VS = 10.
(d) Calculate F0 and hence calculate FL , GL and the load voltage, VL .
(e) Calculate the complex power supplied by the source.
Problem Sheet 7
Page 2 of 2
Ver 8284
E1.1 Analysis of Circuits (2016)
E1.1 Circuit Analysis
Problem Sheet 7 - Solutions
q
The propagation velocity (from page 17-4 of the notes) is u = L01C0 = 2 × 108 m/s which is 23 of
q
L0
the speed of light. The characteristic impedance is Z0 = C
= 100 Ω. These are typical
0
characteristics for the twisted pair cable used in computer networks. The reflection coefficients is
0
given by ρ = R−Z
R+Z0 where R is the Thévenin impedance at the relevant end of the line. This gives
(a) ρS = −0.818, ρL = 0 so a pulse at VS will travel down the line as a forward wave and stop when
it reaches the load (no reflections), (b) ρS = −0.818, ρL = 0.818 so a positive pulse at VS will travel
down the line as a forward wave, be reflected at x = L and travel back towards the source as a
backward wave. When it reaches x = 0 it will be reflected and inverted and will result in a negative
pulse traveling as a forward wave. This whole process will be repeated for ever and you will get an
infinite sequence of pulses each one smaller than than the previous one, (c) ρS = 0, ρL = 0.818 so a
positive pulse at VS will travel down the line as a forward wave, be reflected at x = L and then
travel back towards x = 0 where it will stop. Since the voltage source is DC, f () and g() will be
constants and the waves will be independent of x and t. From Ohm’s law, we know that the
voltage and current in the line are 10 V and 0.2 A so we must have f + g = 10 and
f − g = 0.2 × 100 = 20. Solving these two equations gives f = 15 and g = −5. As we would expect,
the ratio fg = −0.333 is equal to the reflection coefficient at the load. The power carried by the two
waves is
f2
Z0
= 2.25 W and
g2
Z0 = 0.25 W and the difference between these does indeed equal the
−1
102
50 = 2 W. At x = 0, v0 (t) = f (t) + g(t) and i0 (t) = Z0 (f (t) − g(t))
power absorbed by the load:
which gives the waveforms shown inFig. (a). The peak voltage and current
are 9 V and 90 mA
respectively. At x = 300 cm, f t − ux is delayed by 20 ns while g t + ux is advanced by the same
amount. The new voltage and current waveforms are as shown in Fig. (b). The forward and
backward waves now overlap and the peak voltage and current are 12 V and 60 mA respectively.
f(t), g(t) and vL(t) (V)
12
10
vL(t)
8
f(t)
6
4
g(t)
2
0
Fig. (a)
Fig. (b)
0
50
100
Time (s)
150
200
Fig.
0−50
(a) ρL = 100−50
100+50 = 0.33, ρ0 = 0+50 = −1.
(b) In the steady state we must have V = 10 V and I = 0.1 A which means that f + g = 10 and
f − g = 0.1Z0 = 5. We can solve these two equations to give f = 7.5 and g = 2.5.
(c) From the notes (page 17-11) the forward wave is the sum of an infinite number of copies of the
input signal; each extra copy is delayed by an additional round-trip propagation delay ( 600
15 = 40 ns)
and multiplied by an additional factor ρL ρ0 = −0.33. This is shown in Fig. ; the input signal jumps
to 10 V when the switch is closed and stays there forever. Onto this is added a copy of the input
signal delayed by 40 ns and multiplied by −0.33 which means that f (t) jumps to 6.67 at t = 40 ns.
With pulse waveforms, it is sometimes easier to keep track of the changes in the signals rather than
their absolute values. In this case, the input signal only changes once: by +10 V at t = 0. It follows
that f (t) will have a change of +10 at t = 0 followed by a change of 10ρL ρ0 = −3.33 at t = 40
2
3
followed by a change of 10 (ρL ρ0 ) = 1.11 at t = 80 followed by a change of 10 (ρL ρ0 ) = −0.37 at
t = 120 and so on.
So, f (0) = 10, f (40) = 6.67, f (80) = 7.78, f (120) = 7.41, f (160) = 7.53, f (200) = 7.49, f (240) =
7.5), · · ·. As we can see, f (t) → 7.5 as predicted in part (a).
g(t) = ρL f (t − 2L
u ) is just the same as f (t) but delayed by the round-trip propagation delay and
multiplied by ρl = 0.33.
The voltage waveform at x = 0 is equal to f (t) + g(t) and has a constant value of 10 V for t ≥ 0 as
is obvious from the circuit.
The voltage waveform at x = L is vL (t) = fL (t) + gL (t) = f (t − 20) + g(t + 20), that is, the sum of
f (t) delayed by 20 ns and g(t) advanced by 20 ns. This reaches a peak value of 10 (1 + ρL ) = 13.3 V.
Solution Sheet 7
Page 1 of 3
Ver 8284
E1.1 Analysis of Circuits (2016)
Note that the propagation velocity in SI units is 20 × 107 m/s. At 20 MHz the wavenumber is
7
k = ωu = 2π×2×10
= 0.628 rad/m or, equivalently, the wavelength is λ = uf = 2π
20×107
k = 10 m. The
impedance of the line at x = 0, which we will call Z, is, by definition, equal to the phasor Z =
V0
F0 +G0
I0 = Z −1 (F −G ) . Because the line is short circuited at x = L, the reflection coefficient at x = L
0
0
0
is ρL = −1. G0 and F0 are related by G0 = F0 ρL e−2jkL ; the ρL factor comes from the reflection
at x = L and the e−2jkL factor is the phase shift arising from the time delay for the wave to travel
from x = 0 to x = L and back again. Substituting for G0 in the previous expression for Z gives
e−jkL ejkL − e−jkL
1 − e−2jkL
F0 + F0 ρL e−2jkL
= Z0
= Z0 −jkL jkL
Z = Z0
= jZ0 tan kL
F0 − F0 ρL e−2jkL
1 + e−2jkL
e
(e
+ e−jkL )
nλ
1
−1 ZT
which leads to L = nλ
2 + k tan
jZ0 where the 2 term arises because tan() repeats every π. We
1
want to choose L so that (a) ZT = jωC = −159j or (b) ZT = jωL = 126j and in each case choose
n to make L as small as possible while still remaining positive. This gives (a) L = 3.39 m and (b)
L = 1.43 m.
7
π
(a) The wavelength is λ = uf = 20×10
50×106 = 4 m which means that we get a phase shift of − 2 for every
ω
2π
metre we move along the line. Equivalently k = u = λ = 1.571 rad/m. Thus at x = 0, 1, 2, 3, 4, 5,
Fx = F0 e−jkx = 6j, 6, −6j, −6, 6j, 6. Notice that the phase repeats every 4 m since this is one
wavelength.
L −Z0
(b) The reflection coefficient at x = L is ρL = R
RL +Z0 = −0.333.
(c) It follows that GL = ρL FL = −0.333 × 6 = −2. As with Fx in part (a), Gx will get a phase shift
of − π2 for every metre we move backwards along the line towards x = 0, i.e. Gx = GL e−jk(L−x) .
If you substitute for GL you can write this as Gx = ρL FL e−jk(L−x) = ρL F0 e−jkL e−jk(L−x) =
ρL F0 e−jk(2L−x) . Thus at x = 0, 1, 2, 3, 4, 5, Gx = 2j, −2, −2j, 2, 2j, −2.
(d) Vx = Fx + Gx , so at x = 0, 1, 2, 3, 4, 5, Vx = 8j, 4, −8j, −4, 8j, 4. We can see that when Fx
and Gx are in phase (at x = 0, 2, 4) their magnitudes add to give |Vx | = 6 + 2 = 8 whereas when
they are out of phase (at x = 1, 3, 5) the subtract to give |Vx | = 6 − 2 = 4.
1+|ρL |
(e) From part (d), the VSWR is 48 = 2. Because |Gx | = |ρL | |Fx |, the VSWR is 1−|ρ
which, if you
L|
RL Z 0
express ρL in terms of RL equals max Z0 , RL = max (0.5, 2) = 2.
V0
I0
(f) The line impedance at x = 0 is
F0 +G0
= Z0 F
= Z0 8j
4j = Z0 × 2 = 200.
0 −G0
(a) The forward wave is unaffected so Fx = F0 e−jkx = 6j, 6, −6j, −6, 6j, 6 as before.
(b) The line termination is matched and the reflection coefficient is now 0.
(c) It follows that Gx = 0 ∀x.
(d) Vx = Fx + Gx = Fx + 0 = Fx , so Vx = 6j, 6, −6j, −6, 6j, 6.
(e) |Vx | is always equal to 6 and so the VSWR equals 1, its minimum possible value. Measuring the
VSWR gives a way of telling when a line is matched without having to measure either Z0 or RL .
F0 +G0
F0 +0
(f) The line impedance at x = 0 is VI00 = Z0 F
= Z0 F
= Z0 = 100.
0 −G0
0 −0
(a) We have λ = 7.5 m and k = 2π
λ = 0.838. The reflection coefficient is ρL =
G0
−2jkL
= −0.0209 − 0.1989j.
F 0 = ρL e
RL −Z0
RL +Z0
= 0.2. Hence
G
(b) ZT =
V0
I0
F0 +G0
= Z0 F
= Z0
0 −G0
1+ F 0
= 100 × (0.8874 − 0.3677j) = 88.74 − 36.77j = 96.06∠ − 22.5◦ .
0
G
1− F 0
0
T
(c) Hence V0 = VS RSZ+Z
= 10 × (0.911 − 0.033j) = 9.11 − 0.33j.
T
V0
0
(d) Since V0 = F0 + G0 = F0 1 + G
G0 = 9.00 + 1.49j. From this we get
F0 , we can write F0 =
1+ F
−jkL
0
FL = F0 e
= 7.13 − 5.69j. We can now calculate GL = ρL FL = 1.43 − 1.14j and then work out
VL = FL + GL = 8.56 − 6.83j. Notice that once you know F and G at one point on the line, you
can easily calculate their values at any other point whereas V varies in a rather more complicated
manner; because of this, we convert from V to F , then move F to a different place and convert back
to V again.
(e) The complex power supplied by the source is
2
ṼS ∗
(RS + ZT )
Solution Sheet 7
=
1
2
2
|VS |
50
= 0.445 − 0.166j VA
∗ =
98.7 + 36.8j
(RS + ZT )
Page 2 of 3
Ver 8284
E1.1 Analysis of Circuits (2016)
. You might expect this to be purely real because the circuit appears to contain only resistors;
however, there are implicit capacitors and inductors in the transmission line.
Solution Sheet 7
Page 3 of 3